6180 lines
219 KiB
Go
Raw Normal View History

package clickhouseReader
import (
"context"
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
"database/sql"
"encoding/json"
"fmt"
"math"
"math/rand"
2021-05-27 12:52:34 +05:30
"os"
"reflect"
"regexp"
"sort"
2021-05-29 16:32:11 +05:30
"strconv"
"strings"
"sync"
"time"
2021-05-27 12:52:34 +05:30
"go.signoz.io/signoz/pkg/query-service/model/metrics_explorer"
"github.com/go-kit/log"
"github.com/go-kit/log/level"
"github.com/google/uuid"
"github.com/mailru/easyjson"
"github.com/oklog/oklog/pkg/group"
"github.com/pkg/errors"
"github.com/prometheus/common/promlog"
"github.com/prometheus/prometheus/config"
"github.com/prometheus/prometheus/promql"
"github.com/prometheus/prometheus/storage"
"github.com/prometheus/prometheus/storage/remote"
"github.com/prometheus/prometheus/util/stats"
"github.com/ClickHouse/clickhouse-go/v2"
"github.com/ClickHouse/clickhouse-go/v2/lib/driver"
"github.com/jmoiron/sqlx"
"go.signoz.io/signoz/pkg/cache"
"go.signoz.io/signoz/pkg/types/authtypes"
2021-05-27 12:52:34 +05:30
promModel "github.com/prometheus/common/model"
"go.uber.org/zap"
queryprogress "go.signoz.io/signoz/pkg/query-service/app/clickhouseReader/query_progress"
"go.signoz.io/signoz/pkg/query-service/app/logs"
"go.signoz.io/signoz/pkg/query-service/app/resource"
"go.signoz.io/signoz/pkg/query-service/app/services"
"go.signoz.io/signoz/pkg/query-service/app/traces/tracedetail"
"go.signoz.io/signoz/pkg/query-service/common"
"go.signoz.io/signoz/pkg/query-service/constants"
2024-05-15 18:52:01 +05:30
chErrors "go.signoz.io/signoz/pkg/query-service/errors"
am "go.signoz.io/signoz/pkg/query-service/integrations/alertManager"
"go.signoz.io/signoz/pkg/query-service/interfaces"
"go.signoz.io/signoz/pkg/query-service/metrics"
"go.signoz.io/signoz/pkg/query-service/model"
v3 "go.signoz.io/signoz/pkg/query-service/model/v3"
"go.signoz.io/signoz/pkg/query-service/telemetry"
"go.signoz.io/signoz/pkg/query-service/utils"
)
const (
primaryNamespace = "clickhouse"
archiveNamespace = "clickhouse-archive"
signozTraceDBName = "signoz_traces"
signozHistoryDBName = "signoz_analytics"
ruleStateHistoryTableName = "distributed_rule_state_history_v0"
signozDurationMVTable = "distributed_durationSort"
signozUsageExplorerTable = "distributed_usage_explorer"
signozSpansTable = "distributed_signoz_spans"
signozErrorIndexTable = "distributed_signoz_error_index_v2"
signozTraceTableName = "distributed_signoz_index_v2"
signozTraceLocalTableName = "signoz_index_v2"
signozMetricDBName = "signoz_metrics"
signozSampleLocalTableName = "samples_v4"
signozSampleTableName = "distributed_samples_v4"
signozSamplesAgg5mLocalTableName = "samples_v4_agg_5m"
signozSamplesAgg5mTableName = "distributed_samples_v4_agg_5m"
signozSamplesAgg30mLocalTableName = "samples_v4_agg_30m"
signozSamplesAgg30mTableName = "distributed_samples_v4_agg_30m"
signozExpHistLocalTableName = "exp_hist"
signozExpHistTableName = "distributed_exp_hist"
signozTSLocalTableNameV4 = "time_series_v4"
signozTSTableNameV4 = "distributed_time_series_v4"
signozTSLocalTableNameV46Hrs = "time_series_v4_6hrs"
signozTSTableNameV46Hrs = "distributed_time_series_v4_6hrs"
signozTSLocalTableNameV41Day = "time_series_v4_1day"
signozTSTableNameV41Day = "distributed_time_series_v4_1day"
signozTSLocalTableNameV41Week = "time_series_v4_1week"
signozTSTableNameV41Week = "distributed_time_series_v4_1week"
minTimespanForProgressiveSearch = time.Hour
minTimespanForProgressiveSearchMargin = time.Minute
maxProgressiveSteps = 4
charset = "abcdefghijklmnopqrstuvwxyz" +
"ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"
NANOSECOND = 1000000000
)
var (
ErrNoOperationsTable = errors.New("no operations table supplied")
ErrNoIndexTable = errors.New("no index table supplied")
ErrStartTimeRequired = errors.New("start time is required for search queries")
seededRand *rand.Rand = rand.New(
rand.NewSource(time.Now().UnixNano()))
)
// SpanWriter for reading spans from ClickHouse
2021-05-27 12:52:34 +05:30
type ClickHouseReader struct {
db clickhouse.Conn
localDB *sqlx.DB
TraceDB string
operationsTable string
durationTable string
indexTable string
errorTable string
usageExplorerTable string
SpansTable string
spanAttributeTableV2 string
2023-04-25 21:53:46 +05:30
spanAttributesKeysTable string
dependencyGraphTable string
topLevelOperationsTable string
logsDB string
logsTable string
logsLocalTable string
logsAttributeKeys string
logsResourceKeys string
logsTagAttributeTableV2 string
queryEngine *promql.Engine
remoteStorage *remote.Storage
fanoutStorage *storage.Storage
queryProgressTracker queryprogress.QueryProgressTracker
logsTableV2 string
logsLocalTableV2 string
logsResourceTableV2 string
logsResourceLocalTableV2 string
promConfigFile string
promConfig *config.Config
alertManager am.Manager
featureFlags interfaces.FeatureLookup
liveTailRefreshSeconds int
cluster string
2024-09-12 10:58:07 +05:30
useLogsNewSchema bool
useTraceNewSchema bool
logsTableName string
logsLocalTableName string
traceTableName string
traceLocalTableName string
traceResourceTableV3 string
traceSummaryTable string
fluxIntervalForTraceDetail time.Duration
cache cache.Cache
}
// NewTraceReader returns a TraceReader for the database
func NewReader(
localDB *sqlx.DB,
db driver.Conn,
configFile string,
featureFlag interfaces.FeatureLookup,
cluster string,
2024-09-12 10:58:07 +05:30
useLogsNewSchema bool,
useTraceNewSchema bool,
fluxIntervalForTraceDetail time.Duration,
cache cache.Cache,
) *ClickHouseReader {
options := NewOptions(primaryNamespace, archiveNamespace)
return NewReaderFromClickhouseConnection(db, options, localDB, configFile, featureFlag, cluster, useLogsNewSchema, useTraceNewSchema, fluxIntervalForTraceDetail, cache)
}
func NewReaderFromClickhouseConnection(
db driver.Conn,
options *Options,
localDB *sqlx.DB,
configFile string,
featureFlag interfaces.FeatureLookup,
cluster string,
2024-09-12 10:58:07 +05:30
useLogsNewSchema bool,
useTraceNewSchema bool,
fluxIntervalForTraceDetail time.Duration,
cache cache.Cache,
) *ClickHouseReader {
alertManager, err := am.New()
if err != nil {
zap.L().Error("failed to initialize alert manager", zap.Error(err))
zap.L().Error("check if the alert manager URL is correctly set and valid")
os.Exit(1)
}
logsTableName := options.primary.LogsTable
logsLocalTableName := options.primary.LogsLocalTable
if useLogsNewSchema {
logsTableName = options.primary.LogsTableV2
logsLocalTableName = options.primary.LogsLocalTableV2
}
traceTableName := options.primary.IndexTable
traceLocalTableName := options.primary.LocalIndexTable
if useTraceNewSchema {
traceTableName = options.primary.TraceIndexTableV3
traceLocalTableName = options.primary.TraceLocalTableNameV3
}
return &ClickHouseReader{
db: db,
localDB: localDB,
TraceDB: options.primary.TraceDB,
alertManager: alertManager,
operationsTable: options.primary.OperationsTable,
indexTable: options.primary.IndexTable,
errorTable: options.primary.ErrorTable,
usageExplorerTable: options.primary.UsageExplorerTable,
durationTable: options.primary.DurationTable,
SpansTable: options.primary.SpansTable,
spanAttributeTableV2: options.primary.SpanAttributeTableV2,
2023-04-25 21:53:46 +05:30
spanAttributesKeysTable: options.primary.SpanAttributeKeysTable,
dependencyGraphTable: options.primary.DependencyGraphTable,
topLevelOperationsTable: options.primary.TopLevelOperationsTable,
logsDB: options.primary.LogsDB,
logsTable: options.primary.LogsTable,
logsLocalTable: options.primary.LogsLocalTable,
logsAttributeKeys: options.primary.LogsAttributeKeysTable,
logsResourceKeys: options.primary.LogsResourceKeysTable,
logsTagAttributeTableV2: options.primary.LogsTagAttributeTableV2,
liveTailRefreshSeconds: options.primary.LiveTailRefreshSeconds,
promConfigFile: configFile,
featureFlags: featureFlag,
cluster: cluster,
queryProgressTracker: queryprogress.NewQueryProgressTracker(),
useLogsNewSchema: useLogsNewSchema,
useTraceNewSchema: useTraceNewSchema,
logsTableV2: options.primary.LogsTableV2,
logsLocalTableV2: options.primary.LogsLocalTableV2,
logsResourceTableV2: options.primary.LogsResourceTableV2,
logsResourceLocalTableV2: options.primary.LogsResourceLocalTableV2,
logsTableName: logsTableName,
logsLocalTableName: logsLocalTableName,
traceLocalTableName: traceLocalTableName,
traceTableName: traceTableName,
traceResourceTableV3: options.primary.TraceResourceTableV3,
traceSummaryTable: options.primary.TraceSummaryTable,
fluxIntervalForTraceDetail: fluxIntervalForTraceDetail,
cache: cache,
}
}
func (r *ClickHouseReader) Start(readerReady chan bool) {
logLevel := promlog.AllowedLevel{}
logLevel.Set("debug")
2023-03-07 13:37:31 +05:30
allowedFormat := promlog.AllowedFormat{}
allowedFormat.Set("logfmt")
2023-03-07 13:37:31 +05:30
promlogConfig := promlog.Config{
Level: &logLevel,
Format: &allowedFormat,
}
2023-03-07 13:37:31 +05:30
logger := promlog.New(&promlogConfig)
startTime := func() (int64, error) {
return int64(promModel.Latest), nil
}
2023-03-07 13:37:31 +05:30
remoteStorage := remote.NewStorage(
log.With(logger, "component", "remote"),
nil,
startTime,
"",
time.Duration(1*time.Minute),
nil,
2024-10-10 14:10:28 +05:30
false,
2023-03-07 13:37:31 +05:30
)
cfg := struct {
configFile string
localStoragePath string
lookbackDelta promModel.Duration
webTimeout promModel.Duration
queryTimeout promModel.Duration
queryConcurrency int
queryMaxSamples int
RemoteFlushDeadline promModel.Duration
prometheusURL string
logLevel promlog.AllowedLevel
}{
configFile: r.promConfigFile,
}
fanoutStorage := storage.NewFanout(logger, remoteStorage)
opts := promql.EngineOpts{
2023-03-07 13:37:31 +05:30
Logger: log.With(logger, "component", "query engine"),
Reg: nil,
MaxSamples: 50000000,
Timeout: time.Duration(2 * time.Minute),
ActiveQueryTracker: promql.NewActiveQueryTracker(
"",
20,
log.With(logger, "component", "activeQueryTracker"),
),
}
queryEngine := promql.NewEngine(opts)
reloaders := []func(cfg *config.Config) error{
remoteStorage.ApplyConfig,
}
// sync.Once is used to make sure we can close the channel at different execution stages(SIGTERM or when the config is loaded).
type closeOnce struct {
C chan struct{}
once sync.Once
Close func()
}
// Wait until the server is ready to handle reloading.
reloadReady := &closeOnce{
C: make(chan struct{}),
}
reloadReady.Close = func() {
reloadReady.once.Do(func() {
close(reloadReady.C)
})
}
var g group.Group
{
// Initial configuration loading.
cancel := make(chan struct{})
g.Add(
func() error {
var err error
r.promConfig, err = reloadConfig(cfg.configFile, logger, reloaders...)
if err != nil {
return fmt.Errorf("error loading config from %q: %s", cfg.configFile, err)
}
reloadReady.Close()
<-cancel
return nil
},
func(err error) {
close(cancel)
},
)
}
r.queryEngine = queryEngine
r.remoteStorage = remoteStorage
r.fanoutStorage = &fanoutStorage
readerReady <- true
if err := g.Run(); err != nil {
level.Error(logger).Log("err", err)
os.Exit(1)
}
}
func (r *ClickHouseReader) GetQueryEngine() *promql.Engine {
return r.queryEngine
}
func (r *ClickHouseReader) GetFanoutStorage() *storage.Storage {
return r.fanoutStorage
}
func reloadConfig(filename string, logger log.Logger, rls ...func(*config.Config) error) (promConfig *config.Config, err error) {
level.Info(logger).Log("msg", "Loading configuration file", "filename", filename)
2023-03-07 13:37:31 +05:30
conf, err := config.LoadFile(filename, false, false, logger)
if err != nil {
return nil, fmt.Errorf("couldn't load configuration (--config.file=%q): %v", filename, err)
}
failed := false
for _, rl := range rls {
if err := rl(conf); err != nil {
level.Error(logger).Log("msg", "Failed to apply configuration", "err", err)
failed = true
}
}
if failed {
return nil, fmt.Errorf("one or more errors occurred while applying the new configuration (--config.file=%q)", filename)
}
level.Info(logger).Log("msg", "Completed loading of configuration file", "filename", filename)
return conf, nil
}
func (r *ClickHouseReader) GetInstantQueryMetricsResult(ctx context.Context, queryParams *model.InstantQueryMetricsParams) (*promql.Result, *stats.QueryStats, *model.ApiError) {
2023-09-05 18:17:32 +05:30
qry, err := r.queryEngine.NewInstantQuery(ctx, r.remoteStorage, nil, queryParams.Query, queryParams.Time)
if err != nil {
return nil, nil, &model.ApiError{Typ: model.ErrorBadData, Err: err}
}
res := qry.Exec(ctx)
// Optional stats field in response if parameter "stats" is not empty.
2023-03-07 13:37:31 +05:30
var qs stats.QueryStats
if queryParams.Stats != "" {
qs = stats.NewQueryStats(qry.Stats())
}
qry.Close()
2023-03-07 13:37:31 +05:30
return res, &qs, nil
}
func (r *ClickHouseReader) GetQueryRangeResult(ctx context.Context, query *model.QueryRangeParams) (*promql.Result, *stats.QueryStats, *model.ApiError) {
2023-09-05 18:17:32 +05:30
qry, err := r.queryEngine.NewRangeQuery(ctx, r.remoteStorage, nil, query.Query, query.Start, query.End, query.Step)
if err != nil {
return nil, nil, &model.ApiError{Typ: model.ErrorBadData, Err: err}
}
res := qry.Exec(ctx)
// Optional stats field in response if parameter "stats" is not empty.
2023-03-07 13:37:31 +05:30
var qs stats.QueryStats
if query.Stats != "" {
qs = stats.NewQueryStats(qry.Stats())
}
qry.Close()
2023-03-07 13:37:31 +05:30
return res, &qs, nil
}
func (r *ClickHouseReader) GetServicesList(ctx context.Context) (*[]string, error) {
services := []string{}
query := fmt.Sprintf(`SELECT DISTINCT serviceName FROM %s.%s WHERE toDate(timestamp) > now() - INTERVAL 1 DAY`, r.TraceDB, r.traceTableName)
if r.useTraceNewSchema {
query = fmt.Sprintf(`SELECT DISTINCT serviceName FROM %s.%s WHERE ts_bucket_start > (toUnixTimestamp(now() - INTERVAL 1 DAY) - 1800) AND toDate(timestamp) > now() - INTERVAL 1 DAY`, r.TraceDB, r.traceTableName)
}
rows, err := r.db.Query(ctx, query)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
defer rows.Close()
for rows.Next() {
var serviceName string
if err := rows.Scan(&serviceName); err != nil {
return &services, err
}
services = append(services, serviceName)
}
return &services, nil
}
func (r *ClickHouseReader) GetTopLevelOperations(ctx context.Context, skipConfig *model.SkipConfig, start, end time.Time, services []string) (*map[string][]string, *model.ApiError) {
start = start.In(time.UTC)
// The `top_level_operations` that have `time` >= start
operations := map[string][]string{}
// We can't use the `end` because the `top_level_operations` table has the most recent instances of the operations
// We can only use the `start` time to filter the operations
query := fmt.Sprintf(`SELECT name, serviceName, max(time) as ts FROM %s.%s WHERE time >= @start`, r.TraceDB, r.topLevelOperationsTable)
if len(services) > 0 {
query += ` AND serviceName IN @services`
}
query += ` GROUP BY name, serviceName ORDER BY ts DESC LIMIT 5000`
rows, err := r.db.Query(ctx, query, clickhouse.Named("start", start), clickhouse.Named("services", services))
2021-05-31 18:05:54 +05:30
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
defer rows.Close()
for rows.Next() {
var name, serviceName string
var t time.Time
if err := rows.Scan(&name, &serviceName, &t); err != nil {
return nil, &model.ApiError{Typ: model.ErrorInternal, Err: fmt.Errorf("error in reading data")}
}
if _, ok := operations[serviceName]; !ok {
operations[serviceName] = []string{"overflow_operation"}
}
if skipConfig.ShouldSkip(serviceName, name) {
continue
}
operations[serviceName] = append(operations[serviceName], name)
}
return &operations, nil
}
2021-05-31 18:05:54 +05:30
func (r *ClickHouseReader) buildResourceSubQuery(tags []model.TagQueryParam, svc string, start, end time.Time) (string, error) {
// assuming all will be resource attributes.
// and resource attributes are string for traces
filterSet := v3.FilterSet{}
for _, tag := range tags {
// skip the collector id as we don't add it to traces
if tag.Key == "signoz.collector.id" {
continue
}
key := v3.AttributeKey{
Key: tag.Key,
DataType: v3.AttributeKeyDataTypeString,
Type: v3.AttributeKeyTypeResource,
}
it := v3.FilterItem{
Key: key,
}
// as of now only in and not in are supported
switch tag.Operator {
case model.NotInOperator:
it.Operator = v3.FilterOperatorNotIn
it.Value = tag.StringValues
case model.InOperator:
it.Operator = v3.FilterOperatorIn
it.Value = tag.StringValues
default:
return "", fmt.Errorf("operator %s not supported", tag.Operator)
}
filterSet.Items = append(filterSet.Items, it)
}
filterSet.Items = append(filterSet.Items, v3.FilterItem{
Key: v3.AttributeKey{
Key: "service.name",
DataType: v3.AttributeKeyDataTypeString,
Type: v3.AttributeKeyTypeResource,
},
Operator: v3.FilterOperatorEqual,
Value: svc,
})
resourceSubQuery, err := resource.BuildResourceSubQuery(
r.TraceDB,
r.traceResourceTableV3,
start.Unix()-1800,
end.Unix(),
&filterSet,
[]v3.AttributeKey{},
v3.AttributeKey{},
false)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", err
}
return resourceSubQuery, nil
}
func (r *ClickHouseReader) GetServicesV2(ctx context.Context, queryParams *model.GetServicesParams, skipConfig *model.SkipConfig) (*[]model.ServiceItem, *model.ApiError) {
2021-05-31 18:05:54 +05:30
if r.indexTable == "" {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: ErrNoIndexTable}
}
topLevelOps, apiErr := r.GetTopLevelOperations(ctx, skipConfig, *queryParams.Start, *queryParams.End, nil)
if apiErr != nil {
return nil, apiErr
2021-05-31 18:05:54 +05:30
}
serviceItems := []model.ServiceItem{}
var wg sync.WaitGroup
// limit the number of concurrent queries to not overload the clickhouse server
sem := make(chan struct{}, 10)
var mtx sync.RWMutex
for svc, ops := range *topLevelOps {
sem <- struct{}{}
wg.Add(1)
go func(svc string, ops []string) {
defer wg.Done()
defer func() { <-sem }()
var serviceItem model.ServiceItem
var numErrors uint64
// Even if the total number of operations within the time range is less and the all
// the top level operations are high, we want to warn to let user know the issue
// with the instrumentation
serviceItem.DataWarning = model.DataWarning{
TopLevelOps: (*topLevelOps)[svc],
}
// default max_query_size = 262144
// Let's assume the average size of the item in `ops` is 50 bytes
// We can have 262144/50 = 5242 items in the `ops` array
// Although we have make it as big as 5k, We cap the number of items
// in the `ops` array to 1500
ops = ops[:int(math.Min(1500, float64(len(ops))))]
query := fmt.Sprintf(
`SELECT
quantile(0.99)(durationNano) as p99,
avg(durationNano) as avgDuration,
count(*) as numCalls
FROM %s.%s
WHERE serviceName = @serviceName AND name In @names AND timestamp>= @start AND timestamp<= @end`,
r.TraceDB, r.traceTableName,
)
errorQuery := fmt.Sprintf(
`SELECT
count(*) as numErrors
FROM %s.%s
WHERE serviceName = @serviceName AND name In @names AND timestamp>= @start AND timestamp<= @end AND statusCode=2`,
r.TraceDB, r.traceTableName,
)
args := []interface{}{}
args = append(args,
clickhouse.Named("start", strconv.FormatInt(queryParams.Start.UnixNano(), 10)),
clickhouse.Named("end", strconv.FormatInt(queryParams.End.UnixNano(), 10)),
clickhouse.Named("serviceName", svc),
clickhouse.Named("names", ops),
)
resourceSubQuery, err := r.buildResourceSubQuery(queryParams.Tags, svc, *queryParams.Start, *queryParams.End)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return
}
query += `
AND (
resource_fingerprint GLOBAL IN ` +
resourceSubQuery +
`) AND ts_bucket_start >= @start_bucket AND ts_bucket_start <= @end_bucket`
args = append(args,
clickhouse.Named("start_bucket", strconv.FormatInt(queryParams.Start.Unix()-1800, 10)),
clickhouse.Named("end_bucket", strconv.FormatInt(queryParams.End.Unix(), 10)),
)
err = r.db.QueryRow(
ctx,
query,
args...,
).ScanStruct(&serviceItem)
if serviceItem.NumCalls == 0 {
return
}
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return
}
errorQuery += `
AND (
resource_fingerprint GLOBAL IN ` +
resourceSubQuery +
`) AND ts_bucket_start >= @start_bucket AND ts_bucket_start <= @end_bucket`
err = r.db.QueryRow(ctx, errorQuery, args...).Scan(&numErrors)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return
}
serviceItem.ServiceName = svc
serviceItem.NumErrors = numErrors
mtx.Lock()
serviceItems = append(serviceItems, serviceItem)
mtx.Unlock()
}(svc, ops)
}
wg.Wait()
for idx := range serviceItems {
serviceItems[idx].CallRate = float64(serviceItems[idx].NumCalls) / float64(queryParams.Period)
serviceItems[idx].ErrorRate = float64(serviceItems[idx].NumErrors) * 100 / float64(serviceItems[idx].NumCalls)
}
return &serviceItems, nil
}
func (r *ClickHouseReader) GetServices(ctx context.Context, queryParams *model.GetServicesParams, skipConfig *model.SkipConfig) (*[]model.ServiceItem, *model.ApiError) {
if r.useTraceNewSchema {
return r.GetServicesV2(ctx, queryParams, skipConfig)
}
if r.indexTable == "" {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: ErrNoIndexTable}
}
topLevelOps, apiErr := r.GetTopLevelOperations(ctx, skipConfig, *queryParams.Start, *queryParams.End, nil)
if apiErr != nil {
return nil, apiErr
}
serviceItems := []model.ServiceItem{}
var wg sync.WaitGroup
// limit the number of concurrent queries to not overload the clickhouse server
sem := make(chan struct{}, 10)
var mtx sync.RWMutex
for svc, ops := range *topLevelOps {
sem <- struct{}{}
wg.Add(1)
go func(svc string, ops []string) {
defer wg.Done()
defer func() { <-sem }()
var serviceItem model.ServiceItem
var numErrors uint64
// Even if the total number of operations within the time range is less and the all
// the top level operations are high, we want to warn to let user know the issue
// with the instrumentation
serviceItem.DataWarning = model.DataWarning{
TopLevelOps: (*topLevelOps)[svc],
}
// default max_query_size = 262144
// Let's assume the average size of the item in `ops` is 50 bytes
// We can have 262144/50 = 5242 items in the `ops` array
// Although we have make it as big as 5k, We cap the number of items
// in the `ops` array to 1500
ops = ops[:int(math.Min(1500, float64(len(ops))))]
query := fmt.Sprintf(
`SELECT
quantile(0.99)(durationNano) as p99,
avg(durationNano) as avgDuration,
count(*) as numCalls
FROM %s.%s
WHERE serviceName = @serviceName AND name In @names AND timestamp>= @start AND timestamp<= @end`,
r.TraceDB, r.indexTable,
)
errorQuery := fmt.Sprintf(
`SELECT
count(*) as numErrors
FROM %s.%s
WHERE serviceName = @serviceName AND name In @names AND timestamp>= @start AND timestamp<= @end AND statusCode=2`,
r.TraceDB, r.indexTable,
)
args := []interface{}{}
args = append(args,
clickhouse.Named("start", strconv.FormatInt(queryParams.Start.UnixNano(), 10)),
clickhouse.Named("end", strconv.FormatInt(queryParams.End.UnixNano(), 10)),
clickhouse.Named("serviceName", svc),
clickhouse.Named("names", ops),
)
// create TagQuery from TagQueryParams
tags := createTagQueryFromTagQueryParams(queryParams.Tags)
subQuery, argsSubQuery, errStatus := buildQueryWithTagParams(ctx, tags)
query += subQuery
args = append(args, argsSubQuery...)
if errStatus != nil {
zap.L().Error("Error in processing sql query", zap.Error(errStatus))
return
}
err := r.db.QueryRow(
ctx,
query,
args...,
).ScanStruct(&serviceItem)
2021-05-29 16:32:11 +05:30
if serviceItem.NumCalls == 0 {
return
}
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return
}
subQuery, argsSubQuery, errStatus = buildQueryWithTagParams(ctx, tags)
if errStatus != nil {
zap.L().Error("Error building query with tag params", zap.Error(errStatus))
return
}
errorQuery += subQuery
args = append(args, argsSubQuery...)
err = r.db.QueryRow(ctx, errorQuery, args...).Scan(&numErrors)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return
}
2021-05-31 18:05:54 +05:30
serviceItem.ServiceName = svc
serviceItem.NumErrors = numErrors
mtx.Lock()
serviceItems = append(serviceItems, serviceItem)
mtx.Unlock()
}(svc, ops)
2021-05-31 18:05:54 +05:30
}
wg.Wait()
2021-05-31 18:05:54 +05:30
for idx := range serviceItems {
serviceItems[idx].CallRate = float64(serviceItems[idx].NumCalls) / float64(queryParams.Period)
serviceItems[idx].ErrorRate = float64(serviceItems[idx].NumErrors) * 100 / float64(serviceItems[idx].NumCalls)
2021-05-27 12:52:34 +05:30
}
return &serviceItems, nil
}
func getStatusFilters(query string, statusParams []string, excludeMap map[string]struct{}) string {
// status can only be two and if both are selected than they are equivalent to none selected
if _, ok := excludeMap["status"]; ok {
if len(statusParams) == 1 {
if statusParams[0] == "error" {
query += " AND hasError = false"
} else if statusParams[0] == "ok" {
query += " AND hasError = true"
}
}
} else if len(statusParams) == 1 {
if statusParams[0] == "error" {
query += " AND hasError = true"
} else if statusParams[0] == "ok" {
query += " AND hasError = false"
}
}
return query
}
func createTagQueryFromTagQueryParams(queryParams []model.TagQueryParam) []model.TagQuery {
tags := []model.TagQuery{}
for _, tag := range queryParams {
if len(tag.StringValues) > 0 {
tags = append(tags, model.NewTagQueryString(tag))
}
if len(tag.NumberValues) > 0 {
tags = append(tags, model.NewTagQueryNumber(tag))
}
if len(tag.BoolValues) > 0 {
tags = append(tags, model.NewTagQueryBool(tag))
}
}
return tags
}
func StringWithCharset(length int, charset string) string {
b := make([]byte, length)
for i := range b {
b[i] = charset[seededRand.Intn(len(charset))]
}
return string(b)
}
func String(length int) string {
return StringWithCharset(length, charset)
}
func buildQueryWithTagParams(_ context.Context, tags []model.TagQuery) (string, []interface{}, *model.ApiError) {
2023-01-26 01:18:19 +05:30
query := ""
var args []interface{}
for _, item := range tags {
2023-01-26 01:18:19 +05:30
var subQuery string
var argsSubQuery []interface{}
tagMapType := item.GetTagMapColumn()
switch item.GetOperator() {
case model.EqualOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, "=")
case model.NotEqualOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, "!=")
case model.LessThanOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, "<")
case model.GreaterThanOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, ">")
case model.InOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addInOperator(item, tagMapType, false)
case model.NotInOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addInOperator(item, tagMapType, true)
case model.LessThanEqualOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, "<=")
case model.GreaterThanEqualOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addArithmeticOperator(item, tagMapType, ">=")
case model.ContainsOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addContainsOperator(item, tagMapType, false)
case model.NotContainsOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addContainsOperator(item, tagMapType, true)
case model.StartsWithOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addStartsWithOperator(item, tagMapType, false)
case model.NotStartsWithOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addStartsWithOperator(item, tagMapType, true)
case model.ExistsOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addExistsOperator(item, tagMapType, false)
case model.NotExistsOperator:
2023-01-26 01:18:19 +05:30
subQuery, argsSubQuery = addExistsOperator(item, tagMapType, true)
default:
return "", nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("filter operator %s not supported", item.GetOperator())}
}
2023-01-26 01:18:19 +05:30
query += subQuery
args = append(args, argsSubQuery...)
}
2023-01-26 01:18:19 +05:30
return query, args, nil
}
func addInOperator(item model.TagQuery, tagMapType string, not bool) (string, []interface{}) {
values := item.GetValues()
args := []interface{}{}
notStr := ""
if not {
notStr = "NOT"
}
tagValuePair := []string{}
for _, value := range values {
tagKey := "inTagKey" + String(5)
tagValue := "inTagValue" + String(5)
tagValuePair = append(tagValuePair, fmt.Sprintf("%s[@%s] = @%s", tagMapType, tagKey, tagValue))
args = append(args, clickhouse.Named(tagKey, item.GetKey()))
args = append(args, clickhouse.Named(tagValue, value))
}
return fmt.Sprintf(" AND %s (%s)", notStr, strings.Join(tagValuePair, " OR ")), args
}
func addContainsOperator(item model.TagQuery, tagMapType string, not bool) (string, []interface{}) {
values := item.GetValues()
args := []interface{}{}
notStr := ""
if not {
notStr = "NOT"
}
tagValuePair := []string{}
for _, value := range values {
tagKey := "containsTagKey" + String(5)
tagValue := "containsTagValue" + String(5)
tagValuePair = append(tagValuePair, fmt.Sprintf("%s[@%s] ILIKE @%s", tagMapType, tagKey, tagValue))
args = append(args, clickhouse.Named(tagKey, item.GetKey()))
args = append(args, clickhouse.Named(tagValue, "%"+fmt.Sprintf("%v", value)+"%"))
}
return fmt.Sprintf(" AND %s (%s)", notStr, strings.Join(tagValuePair, " OR ")), args
}
func addStartsWithOperator(item model.TagQuery, tagMapType string, not bool) (string, []interface{}) {
values := item.GetValues()
args := []interface{}{}
notStr := ""
if not {
notStr = "NOT"
}
tagValuePair := []string{}
for _, value := range values {
tagKey := "startsWithTagKey" + String(5)
tagValue := "startsWithTagValue" + String(5)
tagValuePair = append(tagValuePair, fmt.Sprintf("%s[@%s] ILIKE @%s", tagMapType, tagKey, tagValue))
args = append(args, clickhouse.Named(tagKey, item.GetKey()))
args = append(args, clickhouse.Named(tagValue, "%"+fmt.Sprintf("%v", value)+"%"))
}
return fmt.Sprintf(" AND %s (%s)", notStr, strings.Join(tagValuePair, " OR ")), args
}
func addArithmeticOperator(item model.TagQuery, tagMapType string, operator string) (string, []interface{}) {
values := item.GetValues()
args := []interface{}{}
tagValuePair := []string{}
for _, value := range values {
tagKey := "arithmeticTagKey" + String(5)
tagValue := "arithmeticTagValue" + String(5)
tagValuePair = append(tagValuePair, fmt.Sprintf("%s[@%s] %s @%s", tagMapType, tagKey, operator, tagValue))
args = append(args, clickhouse.Named(tagKey, item.GetKey()))
args = append(args, clickhouse.Named(tagValue, value))
}
return fmt.Sprintf(" AND (%s)", strings.Join(tagValuePair, " OR ")), args
}
func addExistsOperator(item model.TagQuery, tagMapType string, not bool) (string, []interface{}) {
values := item.GetValues()
notStr := ""
if not {
notStr = "NOT"
}
args := []interface{}{}
tagOperatorPair := []string{}
for range values {
tagKey := "existsTagKey" + String(5)
tagOperatorPair = append(tagOperatorPair, fmt.Sprintf("mapContains(%s, @%s)", tagMapType, tagKey))
args = append(args, clickhouse.Named(tagKey, item.GetKey()))
}
return fmt.Sprintf(" AND %s (%s)", notStr, strings.Join(tagOperatorPair, " OR ")), args
}
func (r *ClickHouseReader) GetTopOperationsV2(ctx context.Context, queryParams *model.GetTopOperationsParams) (*[]model.TopOperationsItem, *model.ApiError) {
namedArgs := []interface{}{
clickhouse.Named("start", strconv.FormatInt(queryParams.Start.UnixNano(), 10)),
clickhouse.Named("end", strconv.FormatInt(queryParams.End.UnixNano(), 10)),
clickhouse.Named("serviceName", queryParams.ServiceName),
clickhouse.Named("start_bucket", strconv.FormatInt(queryParams.Start.Unix()-1800, 10)),
clickhouse.Named("end_bucket", strconv.FormatInt(queryParams.End.Unix(), 10)),
}
var topOperationsItems []model.TopOperationsItem
query := fmt.Sprintf(`
SELECT
quantile(0.5)(durationNano) as p50,
quantile(0.95)(durationNano) as p95,
quantile(0.99)(durationNano) as p99,
COUNT(*) as numCalls,
countIf(statusCode=2) as errorCount,
name
FROM %s.%s
WHERE serviceName = @serviceName AND timestamp>= @start AND timestamp<= @end`,
r.TraceDB, r.traceTableName,
)
resourceSubQuery, err := r.buildResourceSubQuery(queryParams.Tags, queryParams.ServiceName, *queryParams.Start, *queryParams.End)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
query += `
AND (
resource_fingerprint GLOBAL IN ` +
resourceSubQuery +
`) AND ts_bucket_start >= @start_bucket AND ts_bucket_start <= @end_bucket`
query += " GROUP BY name ORDER BY p99 DESC"
if queryParams.Limit > 0 {
query += " LIMIT @limit"
namedArgs = append(namedArgs, clickhouse.Named("limit", queryParams.Limit))
}
err = r.db.Select(ctx, &topOperationsItems, query, namedArgs...)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if topOperationsItems == nil {
topOperationsItems = []model.TopOperationsItem{}
}
return &topOperationsItems, nil
}
func (r *ClickHouseReader) GetTopOperations(ctx context.Context, queryParams *model.GetTopOperationsParams) (*[]model.TopOperationsItem, *model.ApiError) {
if r.useTraceNewSchema {
return r.GetTopOperationsV2(ctx, queryParams)
}
namedArgs := []interface{}{
clickhouse.Named("start", strconv.FormatInt(queryParams.Start.UnixNano(), 10)),
clickhouse.Named("end", strconv.FormatInt(queryParams.End.UnixNano(), 10)),
clickhouse.Named("serviceName", queryParams.ServiceName),
}
var topOperationsItems []model.TopOperationsItem
query := fmt.Sprintf(`
SELECT
quantile(0.5)(durationNano) as p50,
quantile(0.95)(durationNano) as p95,
quantile(0.99)(durationNano) as p99,
COUNT(*) as numCalls,
countIf(statusCode=2) as errorCount,
name
FROM %s.%s
WHERE serviceName = @serviceName AND timestamp>= @start AND timestamp<= @end`,
r.TraceDB, r.indexTable,
)
args := []interface{}{}
args = append(args, namedArgs...)
// create TagQuery from TagQueryParams
tags := createTagQueryFromTagQueryParams(queryParams.Tags)
subQuery, argsSubQuery, errStatus := buildQueryWithTagParams(ctx, tags)
query += subQuery
args = append(args, argsSubQuery...)
if errStatus != nil {
return nil, errStatus
}
query += " GROUP BY name ORDER BY p99 DESC"
if queryParams.Limit > 0 {
query += " LIMIT @limit"
args = append(args, clickhouse.Named("limit", queryParams.Limit))
}
err := r.db.Select(ctx, &topOperationsItems, query, args...)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if topOperationsItems == nil {
topOperationsItems = []model.TopOperationsItem{}
}
return &topOperationsItems, nil
}
func (r *ClickHouseReader) GetUsage(ctx context.Context, queryParams *model.GetUsageParams) (*[]model.UsageItem, error) {
var usageItems []model.UsageItem
namedArgs := []interface{}{
clickhouse.Named("interval", queryParams.StepHour),
clickhouse.Named("start", strconv.FormatInt(queryParams.Start.UnixNano(), 10)),
clickhouse.Named("end", strconv.FormatInt(queryParams.End.UnixNano(), 10)),
}
var query string
if len(queryParams.ServiceName) != 0 {
namedArgs = append(namedArgs, clickhouse.Named("serviceName", queryParams.ServiceName))
query = fmt.Sprintf("SELECT toStartOfInterval(timestamp, INTERVAL @interval HOUR) as time, sum(count) as count FROM %s.%s WHERE service_name=@serviceName AND timestamp>=@start AND timestamp<=@end GROUP BY time ORDER BY time ASC", r.TraceDB, r.usageExplorerTable)
} else {
query = fmt.Sprintf("SELECT toStartOfInterval(timestamp, INTERVAL @interval HOUR) as time, sum(count) as count FROM %s.%s WHERE timestamp>=@start AND timestamp<=@end GROUP BY time ORDER BY time ASC", r.TraceDB, r.usageExplorerTable)
}
err := r.db.Select(ctx, &usageItems, query, namedArgs...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
for i := range usageItems {
usageItems[i].Timestamp = uint64(usageItems[i].Time.UnixNano())
}
if usageItems == nil {
usageItems = []model.UsageItem{}
}
return &usageItems, nil
}
func (r *ClickHouseReader) SearchTracesV2(ctx context.Context, params *model.SearchTracesParams,
smartTraceAlgorithm func(payload []model.SearchSpanResponseItem, targetSpanId string,
levelUp int, levelDown int, spanLimit int) ([]model.SearchSpansResult, error)) (*[]model.SearchSpansResult, error) {
searchSpansResult := []model.SearchSpansResult{
{
Columns: []string{"__time", "SpanId", "TraceId", "ServiceName", "Name", "Kind", "DurationNano", "TagsKeys", "TagsValues", "References", "Events", "HasError", "StatusMessage", "StatusCodeString", "SpanKind"},
IsSubTree: false,
Events: make([][]interface{}, 0),
},
}
var traceSummary model.TraceSummary
summaryQuery := fmt.Sprintf("SELECT * from %s.%s WHERE trace_id=$1", r.TraceDB, r.traceSummaryTable)
err := r.db.QueryRow(ctx, summaryQuery, params.TraceID).Scan(&traceSummary.TraceID, &traceSummary.Start, &traceSummary.End, &traceSummary.NumSpans)
if err != nil {
if err == sql.ErrNoRows {
return &searchSpansResult, nil
}
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
if traceSummary.NumSpans > uint64(params.MaxSpansInTrace) {
zap.L().Error("Max spans allowed in a trace limit reached", zap.Int("MaxSpansInTrace", params.MaxSpansInTrace),
zap.Uint64("Count", traceSummary.NumSpans))
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": traceSummary.NumSpans,
"maxSpansInTraceLimit": params.MaxSpansInTrace,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_MAX_SPANS_ALLOWED_LIMIT_REACHED, data, claims.Email, true, false)
}
return nil, fmt.Errorf("max spans allowed in trace limit reached, please contact support for more details")
}
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": traceSummary.NumSpans,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_TRACE_DETAIL_API, data, claims.Email, true, false)
}
var startTime, endTime, durationNano uint64
var searchScanResponses []model.SpanItemV2
query := fmt.Sprintf("SELECT timestamp, duration_nano, span_id, trace_id, has_error, kind, resource_string_service$$name, name, references, attributes_string, attributes_number, attributes_bool, resources_string, events, status_message, status_code_string, kind_string FROM %s.%s WHERE trace_id=$1 and ts_bucket_start>=$2 and ts_bucket_start<=$3", r.TraceDB, r.traceTableName)
start := time.Now()
err = r.db.Select(ctx, &searchScanResponses, query, params.TraceID, strconv.FormatInt(traceSummary.Start.Unix()-1800, 10), strconv.FormatInt(traceSummary.End.Unix(), 10))
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
end := time.Now()
zap.L().Debug("getTraceSQLQuery took: ", zap.Duration("duration", end.Sub(start)))
searchSpansResult[0].Events = make([][]interface{}, len(searchScanResponses))
searchSpanResponses := []model.SearchSpanResponseItem{}
start = time.Now()
for _, item := range searchScanResponses {
ref := []model.OtelSpanRef{}
err := json.Unmarshal([]byte(item.References), &ref)
if err != nil {
zap.L().Error("Error unmarshalling references", zap.Error(err))
return nil, err
}
// merge attributes_number and attributes_bool to attributes_string
for k, v := range item.Attributes_bool {
item.Attributes_string[k] = fmt.Sprintf("%v", v)
}
for k, v := range item.Attributes_number {
item.Attributes_string[k] = fmt.Sprintf("%v", v)
}
for k, v := range item.Resources_string {
item.Attributes_string[k] = v
}
jsonItem := model.SearchSpanResponseItem{
SpanID: item.SpanID,
TraceID: item.TraceID,
ServiceName: item.ServiceName,
Name: item.Name,
Kind: int32(item.Kind),
DurationNano: int64(item.DurationNano),
HasError: item.HasError,
StatusMessage: item.StatusMessage,
StatusCodeString: item.StatusCodeString,
SpanKind: item.SpanKind,
References: ref,
Events: item.Events,
TagMap: item.Attributes_string,
}
jsonItem.TimeUnixNano = uint64(item.TimeUnixNano.UnixNano() / 1000000)
searchSpanResponses = append(searchSpanResponses, jsonItem)
if startTime == 0 || jsonItem.TimeUnixNano < startTime {
startTime = jsonItem.TimeUnixNano
}
if endTime == 0 || jsonItem.TimeUnixNano > endTime {
endTime = jsonItem.TimeUnixNano
}
if durationNano == 0 || uint64(jsonItem.DurationNano) > durationNano {
durationNano = uint64(jsonItem.DurationNano)
}
}
end = time.Now()
zap.L().Debug("getTraceSQLQuery unmarshal took: ", zap.Duration("duration", end.Sub(start)))
err = r.featureFlags.CheckFeature(model.SmartTraceDetail)
smartAlgoEnabled := err == nil
if len(searchScanResponses) > params.SpansRenderLimit && smartAlgoEnabled {
start = time.Now()
searchSpansResult, err = smartTraceAlgorithm(searchSpanResponses, params.SpanID, params.LevelUp, params.LevelDown, params.SpansRenderLimit)
if err != nil {
return nil, err
}
end = time.Now()
zap.L().Debug("smartTraceAlgo took: ", zap.Duration("duration", end.Sub(start)))
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": len(searchScanResponses),
"spansRenderLimit": params.SpansRenderLimit,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_LARGE_TRACE_OPENED, data, claims.Email, true, false)
}
} else {
for i, item := range searchSpanResponses {
spanEvents := item.GetValues()
searchSpansResult[0].Events[i] = spanEvents
}
}
searchSpansResult[0].StartTimestampMillis = startTime - (durationNano / 1000000)
searchSpansResult[0].EndTimestampMillis = endTime + (durationNano / 1000000)
return &searchSpansResult, nil
}
func (r *ClickHouseReader) SearchTraces(ctx context.Context, params *model.SearchTracesParams,
smartTraceAlgorithm func(payload []model.SearchSpanResponseItem, targetSpanId string,
levelUp int, levelDown int, spanLimit int) ([]model.SearchSpansResult, error)) (*[]model.SearchSpansResult, error) {
if r.useTraceNewSchema {
return r.SearchTracesV2(ctx, params, smartTraceAlgorithm)
}
var countSpans uint64
countQuery := fmt.Sprintf("SELECT count() as count from %s.%s WHERE traceID=$1", r.TraceDB, r.SpansTable)
err := r.db.QueryRow(ctx, countQuery, params.TraceID).Scan(&countSpans)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
if countSpans > uint64(params.MaxSpansInTrace) {
zap.L().Error("Max spans allowed in a trace limit reached", zap.Int("MaxSpansInTrace", params.MaxSpansInTrace),
zap.Uint64("Count", countSpans))
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": countSpans,
"maxSpansInTraceLimit": params.MaxSpansInTrace,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_MAX_SPANS_ALLOWED_LIMIT_REACHED, data, claims.Email, true, false)
}
return nil, fmt.Errorf("max spans allowed in trace limit reached, please contact support for more details")
}
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": countSpans,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_TRACE_DETAIL_API, data, claims.Email, true, false)
}
var startTime, endTime, durationNano uint64
var searchScanResponses []model.SearchSpanDBResponseItem
query := fmt.Sprintf("SELECT timestamp, traceID, model FROM %s.%s WHERE traceID=$1", r.TraceDB, r.SpansTable)
start := time.Now()
err = r.db.Select(ctx, &searchScanResponses, query, params.TraceID)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query")
}
end := time.Now()
zap.L().Debug("getTraceSQLQuery took: ", zap.Duration("duration", end.Sub(start)))
searchSpansResult := []model.SearchSpansResult{{
Columns: []string{"__time", "SpanId", "TraceId", "ServiceName", "Name", "Kind", "DurationNano", "TagsKeys", "TagsValues", "References", "Events", "HasError", "StatusMessage", "StatusCodeString", "SpanKind"},
Events: make([][]interface{}, len(searchScanResponses)),
IsSubTree: false,
},
}
searchSpanResponses := []model.SearchSpanResponseItem{}
start = time.Now()
for _, item := range searchScanResponses {
var jsonItem model.SearchSpanResponseItem
easyjson.Unmarshal([]byte(item.Model), &jsonItem)
jsonItem.TimeUnixNano = uint64(item.Timestamp.UnixNano() / 1000000)
searchSpanResponses = append(searchSpanResponses, jsonItem)
if startTime == 0 || jsonItem.TimeUnixNano < startTime {
startTime = jsonItem.TimeUnixNano
}
if endTime == 0 || jsonItem.TimeUnixNano > endTime {
endTime = jsonItem.TimeUnixNano
}
if durationNano == 0 || uint64(jsonItem.DurationNano) > durationNano {
durationNano = uint64(jsonItem.DurationNano)
}
}
end = time.Now()
zap.L().Debug("getTraceSQLQuery unmarshal took: ", zap.Duration("duration", end.Sub(start)))
err = r.featureFlags.CheckFeature(model.SmartTraceDetail)
smartAlgoEnabled := err == nil
if len(searchScanResponses) > params.SpansRenderLimit && smartAlgoEnabled {
start = time.Now()
searchSpansResult, err = smartTraceAlgorithm(searchSpanResponses, params.SpanID, params.LevelUp, params.LevelDown, params.SpansRenderLimit)
if err != nil {
return nil, err
}
end = time.Now()
zap.L().Debug("smartTraceAlgo took: ", zap.Duration("duration", end.Sub(start)))
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
data := map[string]interface{}{
"traceSize": len(searchScanResponses),
"spansRenderLimit": params.SpansRenderLimit,
}
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_LARGE_TRACE_OPENED, data, claims.Email, true, false)
}
} else {
for i, item := range searchSpanResponses {
spanEvents := item.GetValues()
searchSpansResult[0].Events[i] = spanEvents
}
}
searchSpansResult[0].StartTimestampMillis = startTime - (durationNano / 1000000)
searchSpansResult[0].EndTimestampMillis = endTime + (durationNano / 1000000)
return &searchSpansResult, nil
}
func (r *ClickHouseReader) GetSpansForTrace(ctx context.Context, traceID string, traceDetailsQuery string) ([]model.SpanItemV2, *model.ApiError) {
var traceSummary model.TraceSummary
summaryQuery := fmt.Sprintf("SELECT * from %s.%s WHERE trace_id=$1", r.TraceDB, r.traceSummaryTable)
err := r.db.QueryRow(ctx, summaryQuery, traceID).Scan(&traceSummary.TraceID, &traceSummary.Start, &traceSummary.End, &traceSummary.NumSpans)
if err != nil {
if err == sql.ErrNoRows {
return []model.SpanItemV2{}, nil
}
zap.L().Error("Error in processing trace summary sql query", zap.Error(err))
return nil, model.ExecutionError(fmt.Errorf("error in processing trace summary sql query: %w", err))
}
var searchScanResponses []model.SpanItemV2
queryStartTime := time.Now()
err = r.db.Select(ctx, &searchScanResponses, traceDetailsQuery, traceID, strconv.FormatInt(traceSummary.Start.Unix()-1800, 10), strconv.FormatInt(traceSummary.End.Unix(), 10))
zap.L().Info(traceDetailsQuery)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, model.ExecutionError(fmt.Errorf("error in processing trace data sql query: %w", err))
}
zap.L().Info("trace details query took: ", zap.Duration("duration", time.Since(queryStartTime)), zap.String("traceID", traceID))
return searchScanResponses, nil
}
func (r *ClickHouseReader) GetWaterfallSpansForTraceWithMetadataCache(ctx context.Context, traceID string) (*model.GetWaterfallSpansForTraceWithMetadataCache, error) {
cachedTraceData := new(model.GetWaterfallSpansForTraceWithMetadataCache)
cacheStatus, err := r.cache.Retrieve(ctx, fmt.Sprintf("getWaterfallSpansForTraceWithMetadata-%v", traceID), cachedTraceData, false)
if err != nil {
zap.L().Debug("error in retrieving getWaterfallSpansForTraceWithMetadata cache", zap.Error(err), zap.String("traceID", traceID))
return nil, err
}
if cacheStatus != cache.RetrieveStatusHit {
return nil, errors.Errorf("cache status for getWaterfallSpansForTraceWithMetadata : %s, traceID: %s", cacheStatus, traceID)
}
if time.Since(time.UnixMilli(int64(cachedTraceData.EndTime))) < r.fluxIntervalForTraceDetail {
zap.L().Info("the trace end time falls under the flux interval, skipping getWaterfallSpansForTraceWithMetadata cache", zap.String("traceID", traceID))
return nil, errors.Errorf("the trace end time falls under the flux interval, skipping getWaterfallSpansForTraceWithMetadata cache, traceID: %s", traceID)
}
zap.L().Info("cache is successfully hit, applying cache for getWaterfallSpansForTraceWithMetadata", zap.String("traceID", traceID))
return cachedTraceData, nil
}
func (r *ClickHouseReader) GetWaterfallSpansForTraceWithMetadata(ctx context.Context, traceID string, req *model.GetWaterfallSpansForTraceWithMetadataParams) (*model.GetWaterfallSpansForTraceWithMetadataResponse, *model.ApiError) {
response := new(model.GetWaterfallSpansForTraceWithMetadataResponse)
var startTime, endTime, durationNano, totalErrorSpans, totalSpans uint64
var spanIdToSpanNodeMap = map[string]*model.Span{}
var traceRoots []*model.Span
var serviceNameToTotalDurationMap = map[string]uint64{}
var serviceNameIntervalMap = map[string][]tracedetail.Interval{}
var hasMissingSpans bool
claims, claimsPresent := authtypes.ClaimsFromContext(ctx)
cachedTraceData, err := r.GetWaterfallSpansForTraceWithMetadataCache(ctx, traceID)
if err == nil {
startTime = cachedTraceData.StartTime
endTime = cachedTraceData.EndTime
durationNano = cachedTraceData.DurationNano
spanIdToSpanNodeMap = cachedTraceData.SpanIdToSpanNodeMap
serviceNameToTotalDurationMap = cachedTraceData.ServiceNameToTotalDurationMap
traceRoots = cachedTraceData.TraceRoots
totalSpans = cachedTraceData.TotalSpans
totalErrorSpans = cachedTraceData.TotalErrorSpans
hasMissingSpans = cachedTraceData.HasMissingSpans
if claimsPresent {
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_TRACE_DETAIL_API, map[string]interface{}{"traceSize": totalSpans}, claims.Email, true, false)
}
}
if err != nil {
zap.L().Info("cache miss for getWaterfallSpansForTraceWithMetadata", zap.String("traceID", traceID))
searchScanResponses, err := r.GetSpansForTrace(ctx, traceID, fmt.Sprintf("SELECT timestamp, duration_nano, span_id, trace_id, has_error, kind, resource_string_service$$name, name, references, attributes_string, attributes_number, attributes_bool, resources_string, events, status_message, status_code_string, kind_string FROM %s.%s WHERE trace_id=$1 and ts_bucket_start>=$2 and ts_bucket_start<=$3 ORDER BY timestamp ASC, name ASC", r.TraceDB, r.traceTableName))
if err != nil {
return nil, err
}
if len(searchScanResponses) == 0 {
return response, nil
}
totalSpans = uint64(len(searchScanResponses))
if claimsPresent {
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_TRACE_DETAIL_API, map[string]interface{}{"traceSize": totalSpans}, claims.Email, true, false)
}
processingBeforeCache := time.Now()
for _, item := range searchScanResponses {
ref := []model.OtelSpanRef{}
err := json.Unmarshal([]byte(item.References), &ref)
if err != nil {
zap.L().Error("getWaterfallSpansForTraceWithMetadata: error unmarshalling references", zap.Error(err), zap.String("traceID", traceID))
return nil, model.BadRequest(fmt.Errorf("getWaterfallSpansForTraceWithMetadata: error unmarshalling references %w", err))
}
// merge attributes_number and attributes_bool to attributes_string
for k, v := range item.Attributes_bool {
item.Attributes_string[k] = fmt.Sprintf("%v", v)
}
for k, v := range item.Attributes_number {
item.Attributes_string[k] = fmt.Sprintf("%v", v)
}
for k, v := range item.Resources_string {
item.Attributes_string[k] = v
}
jsonItem := model.Span{
SpanID: item.SpanID,
TraceID: item.TraceID,
ServiceName: item.ServiceName,
Name: item.Name,
Kind: int32(item.Kind),
DurationNano: item.DurationNano,
HasError: item.HasError,
StatusMessage: item.StatusMessage,
StatusCodeString: item.StatusCodeString,
SpanKind: item.SpanKind,
References: ref,
Events: item.Events,
TagMap: item.Attributes_string,
Children: make([]*model.Span, 0),
}
// metadata calculation
startTimeUnixNano := uint64(item.TimeUnixNano.UnixNano())
if startTime == 0 || startTimeUnixNano < startTime {
startTime = startTimeUnixNano
}
if endTime == 0 || (startTimeUnixNano+jsonItem.DurationNano) > endTime {
endTime = (startTimeUnixNano + jsonItem.DurationNano)
}
if durationNano == 0 || jsonItem.DurationNano > durationNano {
durationNano = jsonItem.DurationNano
}
if jsonItem.HasError {
totalErrorSpans = totalErrorSpans + 1
}
// convert start timestamp to millis because right now frontend is expecting it in millis
jsonItem.TimeUnixNano = uint64(item.TimeUnixNano.UnixNano() / 1000000)
// collect the intervals for service for execution time calculation
serviceNameIntervalMap[jsonItem.ServiceName] =
append(serviceNameIntervalMap[jsonItem.ServiceName], tracedetail.Interval{StartTime: jsonItem.TimeUnixNano, Duration: jsonItem.DurationNano / 1000000, Service: jsonItem.ServiceName})
// append to the span node map
spanIdToSpanNodeMap[jsonItem.SpanID] = &jsonItem
}
// traverse through the map and append each node to the children array of the parent node
// and add the missing spans
for _, spanNode := range spanIdToSpanNodeMap {
hasParentSpanNode := false
for _, reference := range spanNode.References {
if reference.RefType == "CHILD_OF" && reference.SpanId != "" {
hasParentSpanNode = true
if parentNode, exists := spanIdToSpanNodeMap[reference.SpanId]; exists {
parentNode.Children = append(parentNode.Children, spanNode)
} else {
// insert the missing span
missingSpan := model.Span{
SpanID: reference.SpanId,
TraceID: spanNode.TraceID,
ServiceName: "",
Name: "Missing Span",
TimeUnixNano: spanNode.TimeUnixNano,
Kind: 0,
DurationNano: spanNode.DurationNano,
HasError: false,
StatusMessage: "",
StatusCodeString: "",
SpanKind: "",
Children: make([]*model.Span, 0),
}
missingSpan.Children = append(missingSpan.Children, spanNode)
spanIdToSpanNodeMap[missingSpan.SpanID] = &missingSpan
traceRoots = append(traceRoots, &missingSpan)
hasMissingSpans = true
}
}
}
if !hasParentSpanNode && !tracedetail.ContainsWaterfallSpan(traceRoots, spanNode) {
traceRoots = append(traceRoots, spanNode)
}
}
// sort the trace roots to add missing spans at the right order
sort.Slice(traceRoots, func(i, j int) bool {
if traceRoots[i].TimeUnixNano == traceRoots[j].TimeUnixNano {
return traceRoots[i].Name < traceRoots[j].Name
}
return traceRoots[i].TimeUnixNano < traceRoots[j].TimeUnixNano
})
serviceNameToTotalDurationMap = tracedetail.CalculateServiceTime(serviceNameIntervalMap)
traceCache := model.GetWaterfallSpansForTraceWithMetadataCache{
StartTime: startTime,
EndTime: endTime,
DurationNano: durationNano,
TotalSpans: totalSpans,
TotalErrorSpans: totalErrorSpans,
SpanIdToSpanNodeMap: spanIdToSpanNodeMap,
ServiceNameToTotalDurationMap: serviceNameToTotalDurationMap,
TraceRoots: traceRoots,
HasMissingSpans: hasMissingSpans,
}
zap.L().Info("getWaterfallSpansForTraceWithMetadata: processing pre cache", zap.Duration("duration", time.Since(processingBeforeCache)), zap.String("traceID", traceID))
cacheErr := r.cache.Store(ctx, fmt.Sprintf("getWaterfallSpansForTraceWithMetadata-%v", traceID), &traceCache, time.Minute*5)
if cacheErr != nil {
zap.L().Debug("failed to store cache for getWaterfallSpansForTraceWithMetadata", zap.String("traceID", traceID), zap.Error(err))
}
}
processingPostCache := time.Now()
selectedSpans, uncollapsedSpans, rootServiceName, rootServiceEntryPoint := tracedetail.GetSelectedSpans(req.UncollapsedSpans, req.SelectedSpanID, traceRoots, spanIdToSpanNodeMap, req.IsSelectedSpanIDUnCollapsed)
zap.L().Info("getWaterfallSpansForTraceWithMetadata: processing post cache", zap.Duration("duration", time.Since(processingPostCache)), zap.String("traceID", traceID))
response.Spans = selectedSpans
response.UncollapsedSpans = uncollapsedSpans
response.StartTimestampMillis = startTime / 1000000
response.EndTimestampMillis = endTime / 1000000
response.TotalSpansCount = totalSpans
response.TotalErrorSpansCount = totalErrorSpans
response.RootServiceName = rootServiceName
response.RootServiceEntryPoint = rootServiceEntryPoint
response.ServiceNameToTotalDurationMap = serviceNameToTotalDurationMap
response.HasMissingSpans = hasMissingSpans
return response, nil
}
func (r *ClickHouseReader) GetFlamegraphSpansForTraceCache(ctx context.Context, traceID string) (*model.GetFlamegraphSpansForTraceCache, error) {
cachedTraceData := new(model.GetFlamegraphSpansForTraceCache)
cacheStatus, err := r.cache.Retrieve(ctx, fmt.Sprintf("getFlamegraphSpansForTrace-%v", traceID), cachedTraceData, false)
if err != nil {
zap.L().Debug("error in retrieving getFlamegraphSpansForTrace cache", zap.Error(err), zap.String("traceID", traceID))
return nil, err
}
if cacheStatus != cache.RetrieveStatusHit {
return nil, errors.Errorf("cache status for getFlamegraphSpansForTrace : %s, traceID: %s", cacheStatus, traceID)
}
if time.Since(time.UnixMilli(int64(cachedTraceData.EndTime))) < r.fluxIntervalForTraceDetail {
zap.L().Info("the trace end time falls under the flux interval, skipping getFlamegraphSpansForTrace cache", zap.String("traceID", traceID))
return nil, errors.Errorf("the trace end time falls under the flux interval, skipping getFlamegraphSpansForTrace cache, traceID: %s", traceID)
}
zap.L().Info("cache is successfully hit, applying cache for getFlamegraphSpansForTrace", zap.String("traceID", traceID))
return cachedTraceData, nil
}
func (r *ClickHouseReader) GetFlamegraphSpansForTrace(ctx context.Context, traceID string, req *model.GetFlamegraphSpansForTraceParams) (*model.GetFlamegraphSpansForTraceResponse, *model.ApiError) {
trace := new(model.GetFlamegraphSpansForTraceResponse)
var startTime, endTime, durationNano uint64
var spanIdToSpanNodeMap = map[string]*model.FlamegraphSpan{}
// map[traceID][level]span
var selectedSpans = [][]*model.FlamegraphSpan{}
var traceRoots []*model.FlamegraphSpan
// get the trace tree from cache!
cachedTraceData, err := r.GetFlamegraphSpansForTraceCache(ctx, traceID)
if err == nil {
startTime = cachedTraceData.StartTime
endTime = cachedTraceData.EndTime
durationNano = cachedTraceData.DurationNano
selectedSpans = cachedTraceData.SelectedSpans
traceRoots = cachedTraceData.TraceRoots
}
if err != nil {
zap.L().Info("cache miss for getFlamegraphSpansForTrace", zap.String("traceID", traceID))
searchScanResponses, err := r.GetSpansForTrace(ctx, traceID, fmt.Sprintf("SELECT timestamp, duration_nano, span_id, trace_id, has_error,references, resource_string_service$$name, name FROM %s.%s WHERE trace_id=$1 and ts_bucket_start>=$2 and ts_bucket_start<=$3 ORDER BY timestamp ASC, name ASC", r.TraceDB, r.traceTableName))
if err != nil {
return nil, err
}
if len(searchScanResponses) == 0 {
return trace, nil
}
processingBeforeCache := time.Now()
for _, item := range searchScanResponses {
ref := []model.OtelSpanRef{}
err := json.Unmarshal([]byte(item.References), &ref)
if err != nil {
zap.L().Error("Error unmarshalling references", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorBadData, Err: fmt.Errorf("error in unmarshalling references: %w", err)}
}
jsonItem := model.FlamegraphSpan{
SpanID: item.SpanID,
TraceID: item.TraceID,
ServiceName: item.ServiceName,
Name: item.Name,
DurationNano: item.DurationNano,
HasError: item.HasError,
References: ref,
Children: make([]*model.FlamegraphSpan, 0),
}
// metadata calculation
startTimeUnixNano := uint64(item.TimeUnixNano.UnixNano())
if startTime == 0 || startTimeUnixNano < startTime {
startTime = startTimeUnixNano
}
if endTime == 0 || (startTimeUnixNano+jsonItem.DurationNano) > endTime {
endTime = (startTimeUnixNano + jsonItem.DurationNano)
}
if durationNano == 0 || jsonItem.DurationNano > durationNano {
durationNano = jsonItem.DurationNano
}
jsonItem.TimeUnixNano = uint64(item.TimeUnixNano.UnixNano() / 1000000)
spanIdToSpanNodeMap[jsonItem.SpanID] = &jsonItem
}
// traverse through the map and append each node to the children array of the parent node
// and add missing spans
for _, spanNode := range spanIdToSpanNodeMap {
hasParentSpanNode := false
for _, reference := range spanNode.References {
if reference.RefType == "CHILD_OF" && reference.SpanId != "" {
hasParentSpanNode = true
if parentNode, exists := spanIdToSpanNodeMap[reference.SpanId]; exists {
parentNode.Children = append(parentNode.Children, spanNode)
} else {
// insert the missing spans
missingSpan := model.FlamegraphSpan{
SpanID: reference.SpanId,
TraceID: spanNode.TraceID,
ServiceName: "",
Name: "Missing Span",
TimeUnixNano: spanNode.TimeUnixNano,
DurationNano: spanNode.DurationNano,
HasError: false,
Children: make([]*model.FlamegraphSpan, 0),
}
missingSpan.Children = append(missingSpan.Children, spanNode)
spanIdToSpanNodeMap[missingSpan.SpanID] = &missingSpan
traceRoots = append(traceRoots, &missingSpan)
}
}
}
if !hasParentSpanNode && !tracedetail.ContainsFlamegraphSpan(traceRoots, spanNode) {
traceRoots = append(traceRoots, spanNode)
}
}
selectedSpans = tracedetail.GetSelectedSpansForFlamegraph(traceRoots, spanIdToSpanNodeMap)
traceCache := model.GetFlamegraphSpansForTraceCache{
StartTime: startTime,
EndTime: endTime,
DurationNano: durationNano,
SelectedSpans: selectedSpans,
TraceRoots: traceRoots,
}
zap.L().Info("getFlamegraphSpansForTrace: processing pre cache", zap.Duration("duration", time.Since(processingBeforeCache)), zap.String("traceID", traceID))
cacheErr := r.cache.Store(ctx, fmt.Sprintf("getFlamegraphSpansForTrace-%v", traceID), &traceCache, time.Minute*5)
if cacheErr != nil {
zap.L().Debug("failed to store cache for getFlamegraphSpansForTrace", zap.String("traceID", traceID), zap.Error(err))
}
}
processingPostCache := time.Now()
feat(trace-details): frontend changes for trace details (#6905) * feat(trace-details): frontend changes for trace details * feat(trace-detail): address review comments from elipsis * feat(trace0-detail): add the new drawer designs * feat(trace-detail): handle the selected span hover * feat(trace-detail): address theme colors and span selection * feat(trace-detail): fix some more css * feat(trace-detail): fix some more css * feat(trace-detail): add hoverred span and handled no data components for new drawer * feat(trace-detail): handle light mode designs * feat(trace-detail): remove the hover functionality in favor of performance * feat(trace-detail): span lines connectors * feat(trace-detail): span lines connectors * feat(trace-detail): handle the line matching for flamegraph and waterfall * feat(trace-waterfall): change the timeline color to make it less poky * feat(trace-waterfall): added where clause support in trace details page * feat(trace-waterfall): added where clause support in trace details page * feat(trace-detail): handle light mode designs * feat(trace-detail): handle light mode designs * feat(trace-detail): fix build issues * feat(trace-detail): handle loading error state for filters and flamegraph hovered state * feat(trace-detail): fix the hardcoded traceID * feat(trace-detail): remove unnecessaru use effects * feat(trace-detail): handled the flamegraph update with ID * feat(trace-detail): added timestamp bucketing and latency sampling * feat(trace-detail): extract the buckets and span limit in constants * feat(trace-detail): minor VQA comments * feat(trace-detail): remove unnecessaru use effects * feat(trace-detail): add go to related logs * feat(trace-detail): address review comments * feat(trace-detail): address review comments * feat(trace-detail): address review comments * feat(trace-detail): address review comments
2025-01-28 22:04:24 +05:30
selectedSpansForRequest := tracedetail.GetSelectedSpansForFlamegraphForRequest(req.SelectedSpanID, selectedSpans, startTime, endTime)
zap.L().Info("getFlamegraphSpansForTrace: processing post cache", zap.Duration("duration", time.Since(processingPostCache)), zap.String("traceID", traceID))
trace.Spans = selectedSpansForRequest
trace.StartTimestampMillis = startTime / 1000000
trace.EndTimestampMillis = endTime / 1000000
return trace, nil
}
func (r *ClickHouseReader) GetDependencyGraph(ctx context.Context, queryParams *model.GetServicesParams) (*[]model.ServiceMapDependencyResponseItem, error) {
response := []model.ServiceMapDependencyResponseItem{}
args := []interface{}{}
args = append(args,
clickhouse.Named("start", uint64(queryParams.Start.Unix())),
clickhouse.Named("end", uint64(queryParams.End.Unix())),
clickhouse.Named("duration", uint64(queryParams.End.Unix()-queryParams.Start.Unix())),
)
query := fmt.Sprintf(`
WITH
quantilesMergeState(0.5, 0.75, 0.9, 0.95, 0.99)(duration_quantiles_state) AS duration_quantiles_state,
finalizeAggregation(duration_quantiles_state) AS result
SELECT
src as parent,
dest as child,
result[1] AS p50,
result[2] AS p75,
result[3] AS p90,
result[4] AS p95,
result[5] AS p99,
sum(total_count) as callCount,
sum(total_count)/ @duration AS callRate,
sum(error_count)/sum(total_count) * 100 as errorRate
FROM %s.%s
WHERE toUInt64(toDateTime(timestamp)) >= @start AND toUInt64(toDateTime(timestamp)) <= @end`,
r.TraceDB, r.dependencyGraphTable,
)
tags := createTagQueryFromTagQueryParams(queryParams.Tags)
filterQuery, filterArgs := services.BuildServiceMapQuery(tags)
query += filterQuery + " GROUP BY src, dest;"
args = append(args, filterArgs...)
zap.L().Debug("GetDependencyGraph query", zap.String("query", query), zap.Any("args", args))
err := r.db.Select(ctx, &response, query, args...)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, fmt.Errorf("error in processing sql query %w", err)
}
return &response, nil
}
2021-06-01 15:13:48 +05:30
func getLocalTableName(tableName string) string {
tableNameSplit := strings.Split(tableName, ".")
return tableNameSplit[0] + "." + strings.Split(tableNameSplit[1], "distributed_")[1]
}
func (r *ClickHouseReader) SetTTLLogsV2(ctx context.Context, params *model.TTLParams) (*model.SetTTLResponseItem, *model.ApiError) {
// Keep only latest 100 transactions/requests
r.deleteTtlTransactions(ctx, 100)
// uuid is used as transaction id
uuidWithHyphen := uuid.New()
uuid := strings.Replace(uuidWithHyphen.String(), "-", "", -1)
coldStorageDuration := -1
if len(params.ColdStorageVolume) > 0 {
coldStorageDuration = int(params.ToColdStorageDuration)
}
tableNameArray := []string{r.logsDB + "." + r.logsLocalTableV2, r.logsDB + "." + r.logsResourceLocalTableV2}
// check if there is existing things to be done
for _, tableName := range tableNameArray {
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err != nil {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending {
return nil, &model.ApiError{Typ: model.ErrorConflict, Err: fmt.Errorf("TTL is already running")}
}
}
// TTL query for logs_v2 table
ttlLogsV2 := fmt.Sprintf(
"ALTER TABLE %v ON CLUSTER %s MODIFY TTL toDateTime(timestamp / 1000000000) + "+
"INTERVAL %v SECOND DELETE", tableNameArray[0], r.cluster, params.DelDuration)
if len(params.ColdStorageVolume) > 0 {
ttlLogsV2 += fmt.Sprintf(", toDateTime(timestamp / 1000000000)"+
" + INTERVAL %v SECOND TO VOLUME '%s'",
params.ToColdStorageDuration, params.ColdStorageVolume)
}
// TTL query for logs_v2_resource table
// adding 1800 as our bucket size is 1800 seconds
ttlLogsV2Resource := fmt.Sprintf(
"ALTER TABLE %v ON CLUSTER %s MODIFY TTL toDateTime(seen_at_ts_bucket_start) + toIntervalSecond(1800) + "+
"INTERVAL %v SECOND DELETE", tableNameArray[1], r.cluster, params.DelDuration)
if len(params.ColdStorageVolume) > 0 {
ttlLogsV2Resource += fmt.Sprintf(", toDateTime(seen_at_ts_bucket_start) + toIntervalSecond(1800) + "+
"INTERVAL %v SECOND TO VOLUME '%s'",
params.ToColdStorageDuration, params.ColdStorageVolume)
}
ttlPayload := map[string]string{
tableNameArray[0]: ttlLogsV2,
tableNameArray[1]: ttlLogsV2Resource,
}
// set the ttl if nothing is pending/ no errors
go func(ttlPayload map[string]string) {
for tableName, query := range ttlPayload {
// https://github.com/SigNoz/signoz/issues/5470
// we will change ttl for only the new parts and not the old ones
query += " SETTINGS materialize_ttl_after_modify=0"
_, dbErr := r.localDB.Exec("INSERT INTO ttl_status (transaction_id, created_at, updated_at, table_name, ttl, status, cold_storage_ttl) VALUES (?, ?, ?, ?, ?, ?, ?)", uuid, time.Now(), time.Now(), tableName, params.DelDuration, constants.StatusPending, coldStorageDuration)
if dbErr != nil {
zap.L().Error("error in inserting to ttl_status table", zap.Error(dbErr))
return
}
err := r.setColdStorage(context.Background(), tableName, params.ColdStorageVolume)
if err != nil {
zap.L().Error("error in setting cold storage", zap.Error(err))
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err == nil {
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
return
}
zap.L().Info("Executing TTL request: ", zap.String("request", query))
statusItem, _ := r.checkTTLStatusItem(ctx, tableName)
if err := r.db.Exec(ctx, query); err != nil {
zap.L().Error("error while setting ttl", zap.Error(err))
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
return
}
_, dbErr = r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusSuccess, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
}(ttlPayload)
return &model.SetTTLResponseItem{Message: "move ttl has been successfully set up"}, nil
}
func (r *ClickHouseReader) SetTTLTracesV2(ctx context.Context, params *model.TTLParams) (*model.SetTTLResponseItem, *model.ApiError) {
// uuid is used as transaction id
uuidWithHyphen := uuid.New()
uuid := strings.Replace(uuidWithHyphen.String(), "-", "", -1)
tableNames := []string{
r.TraceDB + "." + r.traceTableName,
r.TraceDB + "." + r.traceResourceTableV3,
r.TraceDB + "." + signozErrorIndexTable,
r.TraceDB + "." + signozUsageExplorerTable,
r.TraceDB + "." + defaultDependencyGraphTable,
r.TraceDB + "." + r.traceSummaryTable,
}
coldStorageDuration := -1
if len(params.ColdStorageVolume) > 0 {
coldStorageDuration = int(params.ToColdStorageDuration)
}
// check if there is existing things to be done
for _, tableName := range tableNames {
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err != nil {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending {
return nil, &model.ApiError{Typ: model.ErrorConflict, Err: fmt.Errorf("TTL is already running")}
}
}
// TTL query
ttlV2 := "ALTER TABLE %s ON CLUSTER %s MODIFY TTL toDateTime(%s) + INTERVAL %v SECOND DELETE"
ttlV2ColdStorage := ", toDateTime(%s) + INTERVAL %v SECOND TO VOLUME '%s'"
// TTL query for resource table
ttlV2Resource := "ALTER TABLE %s ON CLUSTER %s MODIFY TTL toDateTime(seen_at_ts_bucket_start) + toIntervalSecond(1800) + INTERVAL %v SECOND DELETE"
ttlTracesV2ResourceColdStorage := ", toDateTime(seen_at_ts_bucket_start) + toIntervalSecond(1800) + INTERVAL %v SECOND TO VOLUME '%s'"
for _, distributedTableName := range tableNames {
go func(distributedTableName string) {
tableName := getLocalTableName(distributedTableName)
// for trace summary table, we need to use end instead of timestamp
timestamp := "timestamp"
if strings.HasSuffix(distributedTableName, r.traceSummaryTable) {
timestamp = "end"
}
_, dbErr := r.localDB.Exec("INSERT INTO ttl_status (transaction_id, created_at, updated_at, table_name, ttl, status, cold_storage_ttl) VALUES (?, ?, ?, ?, ?, ?, ?)", uuid, time.Now(), time.Now(), tableName, params.DelDuration, constants.StatusPending, coldStorageDuration)
if dbErr != nil {
zap.L().Error("Error in inserting to ttl_status table", zap.Error(dbErr))
return
}
req := fmt.Sprintf(ttlV2, tableName, r.cluster, timestamp, params.DelDuration)
if strings.HasSuffix(distributedTableName, r.traceResourceTableV3) {
req = fmt.Sprintf(ttlV2Resource, tableName, r.cluster, params.DelDuration)
}
if len(params.ColdStorageVolume) > 0 {
if strings.HasSuffix(distributedTableName, r.traceResourceTableV3) {
req += fmt.Sprintf(ttlTracesV2ResourceColdStorage, params.ToColdStorageDuration, params.ColdStorageVolume)
} else {
req += fmt.Sprintf(ttlV2ColdStorage, timestamp, params.ToColdStorageDuration, params.ColdStorageVolume)
}
}
err := r.setColdStorage(context.Background(), tableName, params.ColdStorageVolume)
if err != nil {
zap.L().Error("Error in setting cold storage", zap.Error(err))
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err == nil {
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
return
}
req += " SETTINGS materialize_ttl_after_modify=0;"
zap.L().Error(" ExecutingTTL request: ", zap.String("request", req))
statusItem, _ := r.checkTTLStatusItem(ctx, tableName)
if err := r.db.Exec(ctx, req); err != nil {
zap.L().Error("Error in executing set TTL query", zap.Error(err))
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
return
}
_, dbErr = r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusSuccess, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}(distributedTableName)
}
return &model.SetTTLResponseItem{Message: "move ttl has been successfully set up"}, nil
}
// SetTTL sets the TTL for traces or metrics or logs tables.
// This is an async API which creates goroutines to set TTL.
// Status of TTL update is tracked with ttl_status table in sqlite db.
2022-08-04 14:28:10 +05:30
// SetTTL sets the TTL for traces or metrics or logs tables.
// This is an async API which creates goroutines to set TTL.
// Status of TTL update is tracked with ttl_status table in sqlite db.
func (r *ClickHouseReader) SetTTL(ctx context.Context,
params *model.TTLParams) (*model.SetTTLResponseItem, *model.ApiError) {
// Keep only latest 100 transactions/requests
r.deleteTtlTransactions(ctx, 100)
// uuid is used as transaction id
uuidWithHyphen := uuid.New()
uuid := strings.Replace(uuidWithHyphen.String(), "-", "", -1)
coldStorageDuration := -1
if len(params.ColdStorageVolume) > 0 {
coldStorageDuration = int(params.ToColdStorageDuration)
}
switch params.Type {
case constants.TraceTTL:
if r.useTraceNewSchema {
return r.SetTTLTracesV2(ctx, params)
}
tableNames := []string{
signozTraceDBName + "." + signozTraceTableName,
signozTraceDBName + "." + signozDurationMVTable,
signozTraceDBName + "." + signozSpansTable,
signozTraceDBName + "." + signozErrorIndexTable,
signozTraceDBName + "." + signozUsageExplorerTable,
signozTraceDBName + "." + defaultDependencyGraphTable,
}
for _, tableName := range tableNames {
tableName := getLocalTableName(tableName)
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err != nil {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending {
return nil, &model.ApiError{Typ: model.ErrorConflict, Err: fmt.Errorf("TTL is already running")}
}
}
for _, tableName := range tableNames {
tableName := getLocalTableName(tableName)
// TODO: DB queries should be implemented with transactional statements but currently clickhouse doesn't support them. Issue: https://github.com/ClickHouse/ClickHouse/issues/22086
go func(tableName string) {
_, dbErr := r.localDB.Exec("INSERT INTO ttl_status (transaction_id, created_at, updated_at, table_name, ttl, status, cold_storage_ttl) VALUES (?, ?, ?, ?, ?, ?, ?)", uuid, time.Now(), time.Now(), tableName, params.DelDuration, constants.StatusPending, coldStorageDuration)
if dbErr != nil {
zap.L().Error("Error in inserting to ttl_status table", zap.Error(dbErr))
return
}
req := fmt.Sprintf(
"ALTER TABLE %v ON CLUSTER %s MODIFY TTL toDateTime(timestamp) + INTERVAL %v SECOND DELETE",
tableName, r.cluster, params.DelDuration)
if len(params.ColdStorageVolume) > 0 {
req += fmt.Sprintf(", toDateTime(timestamp) + INTERVAL %v SECOND TO VOLUME '%s'",
params.ToColdStorageDuration, params.ColdStorageVolume)
}
err := r.setColdStorage(context.Background(), tableName, params.ColdStorageVolume)
if err != nil {
zap.L().Error("Error in setting cold storage", zap.Error(err))
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err == nil {
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
return
}
req += " SETTINGS materialize_ttl_after_modify=0;"
zap.L().Error("Executing TTL request: ", zap.String("request", req))
statusItem, _ := r.checkTTLStatusItem(ctx, tableName)
if err := r.db.Exec(context.Background(), req); err != nil {
zap.L().Error("Error in executing set TTL query", zap.Error(err))
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
return
}
_, dbErr = r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusSuccess, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}(tableName)
}
case constants.MetricsTTL:
tableNames := []string{
signozMetricDBName + "." + signozSampleLocalTableName,
signozMetricDBName + "." + signozSamplesAgg5mLocalTableName,
signozMetricDBName + "." + signozSamplesAgg30mLocalTableName,
signozMetricDBName + "." + signozExpHistLocalTableName,
signozMetricDBName + "." + signozTSLocalTableNameV4,
signozMetricDBName + "." + signozTSLocalTableNameV46Hrs,
signozMetricDBName + "." + signozTSLocalTableNameV41Day,
signozMetricDBName + "." + signozTSLocalTableNameV41Week,
}
for _, tableName := range tableNames {
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err != nil {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending {
return nil, &model.ApiError{Typ: model.ErrorConflict, Err: fmt.Errorf("TTL is already running")}
}
}
metricTTL := func(tableName string) {
_, dbErr := r.localDB.Exec("INSERT INTO ttl_status (transaction_id, created_at, updated_at, table_name, ttl, status, cold_storage_ttl) VALUES (?, ?, ?, ?, ?, ?, ?)", uuid, time.Now(), time.Now(), tableName, params.DelDuration, constants.StatusPending, coldStorageDuration)
if dbErr != nil {
zap.L().Error("Error in inserting to ttl_status table", zap.Error(dbErr))
return
}
timeColumn := "timestamp_ms"
if strings.Contains(tableName, "v4") || strings.Contains(tableName, "exp_hist") {
timeColumn = "unix_milli"
}
req := fmt.Sprintf(
"ALTER TABLE %v ON CLUSTER %s MODIFY TTL toDateTime(toUInt32(%s / 1000), 'UTC') + "+
"INTERVAL %v SECOND DELETE", tableName, r.cluster, timeColumn, params.DelDuration)
if len(params.ColdStorageVolume) > 0 {
req += fmt.Sprintf(", toDateTime(toUInt32(%s / 1000), 'UTC')"+
" + INTERVAL %v SECOND TO VOLUME '%s'",
timeColumn, params.ToColdStorageDuration, params.ColdStorageVolume)
}
err := r.setColdStorage(context.Background(), tableName, params.ColdStorageVolume)
if err != nil {
zap.L().Error("Error in setting cold storage", zap.Error(err))
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err == nil {
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
return
}
req += " SETTINGS materialize_ttl_after_modify=0"
zap.L().Info("Executing TTL request: ", zap.String("request", req))
statusItem, _ := r.checkTTLStatusItem(ctx, tableName)
if err := r.db.Exec(ctx, req); err != nil {
zap.L().Error("error while setting ttl.", zap.Error(err))
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
return
}
_, dbErr = r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusSuccess, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
return
}
}
for _, tableName := range tableNames {
go metricTTL(tableName)
}
2022-08-04 14:28:10 +05:30
case constants.LogsTTL:
if r.useLogsNewSchema {
return r.SetTTLLogsV2(ctx, params)
}
tableName := r.logsDB + "." + r.logsLocalTable
2022-08-04 14:28:10 +05:30
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err != nil {
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending {
return nil, &model.ApiError{Typ: model.ErrorConflict, Err: fmt.Errorf("TTL is already running")}
}
go func(tableName string) {
_, dbErr := r.localDB.Exec("INSERT INTO ttl_status (transaction_id, created_at, updated_at, table_name, ttl, status, cold_storage_ttl) VALUES (?, ?, ?, ?, ?, ?, ?)", uuid, time.Now(), time.Now(), tableName, params.DelDuration, constants.StatusPending, coldStorageDuration)
if dbErr != nil {
zap.L().Error("error in inserting to ttl_status table", zap.Error(dbErr))
2022-08-04 14:28:10 +05:30
return
}
req := fmt.Sprintf(
"ALTER TABLE %v ON CLUSTER %s MODIFY TTL toDateTime(timestamp / 1000000000) + "+
"INTERVAL %v SECOND DELETE", tableName, r.cluster, params.DelDuration)
2022-08-04 14:28:10 +05:30
if len(params.ColdStorageVolume) > 0 {
req += fmt.Sprintf(", toDateTime(timestamp / 1000000000)"+
" + INTERVAL %v SECOND TO VOLUME '%s'",
params.ToColdStorageDuration, params.ColdStorageVolume)
}
err := r.setColdStorage(context.Background(), tableName, params.ColdStorageVolume)
if err != nil {
zap.L().Error("error in setting cold storage", zap.Error(err))
2022-08-04 14:28:10 +05:30
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
if err == nil {
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
2022-08-04 14:28:10 +05:30
return
}
}
return
}
req += " SETTINGS materialize_ttl_after_modify=0"
zap.L().Info("Executing TTL request: ", zap.String("request", req))
2022-08-04 14:28:10 +05:30
statusItem, _ := r.checkTTLStatusItem(ctx, tableName)
if err := r.db.Exec(ctx, req); err != nil {
zap.L().Error("error while setting ttl", zap.Error(err))
2022-08-04 14:28:10 +05:30
_, dbErr := r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusFailed, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
2022-08-04 14:28:10 +05:30
return
}
return
}
_, dbErr = r.localDB.Exec("UPDATE ttl_status SET updated_at = ?, status = ? WHERE id = ?", time.Now(), constants.StatusSuccess, statusItem.Id)
if dbErr != nil {
zap.L().Error("Error in processing ttl_status update sql query", zap.Error(dbErr))
2022-08-04 14:28:10 +05:30
return
}
}(tableName)
default:
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while setting ttl. ttl type should be <metrics|traces>, got %v",
params.Type)}
}
return &model.SetTTLResponseItem{Message: "move ttl has been successfully set up"}, nil
}
func (r *ClickHouseReader) deleteTtlTransactions(_ context.Context, numberOfTransactionsStore int) {
_, err := r.localDB.Exec("DELETE FROM ttl_status WHERE transaction_id NOT IN (SELECT distinct transaction_id FROM ttl_status ORDER BY created_at DESC LIMIT ?)", numberOfTransactionsStore)
if err != nil {
zap.L().Error("Error in processing ttl_status delete sql query", zap.Error(err))
}
}
// checkTTLStatusItem checks if ttl_status table has an entry for the given table name
func (r *ClickHouseReader) checkTTLStatusItem(_ context.Context, tableName string) (model.TTLStatusItem, *model.ApiError) {
statusItem := []model.TTLStatusItem{}
query := `SELECT id, status, ttl, cold_storage_ttl FROM ttl_status WHERE table_name = ? ORDER BY created_at DESC`
zap.L().Info("checkTTLStatusItem query", zap.String("query", query), zap.String("tableName", tableName))
stmt, err := r.localDB.Preparex(query)
if err != nil {
zap.L().Error("Error preparing query for checkTTLStatusItem", zap.Error(err))
return model.TTLStatusItem{}, &model.ApiError{Typ: model.ErrorInternal, Err: err}
}
err = stmt.Select(&statusItem, tableName)
if len(statusItem) == 0 {
return model.TTLStatusItem{}, nil
}
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return model.TTLStatusItem{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
return statusItem[0], nil
}
// setTTLQueryStatus fetches ttl_status table status from DB
func (r *ClickHouseReader) setTTLQueryStatus(ctx context.Context, tableNameArray []string) (string, *model.ApiError) {
failFlag := false
status := constants.StatusSuccess
for _, tableName := range tableNameArray {
statusItem, err := r.checkTTLStatusItem(ctx, tableName)
emptyStatusStruct := model.TTLStatusItem{}
if statusItem == emptyStatusStruct {
return "", nil
}
if err != nil {
return "", &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing ttl_status check sql query")}
}
if statusItem.Status == constants.StatusPending && statusItem.UpdatedAt.Unix()-time.Now().Unix() < 3600 {
status = constants.StatusPending
return status, nil
}
if statusItem.Status == constants.StatusFailed {
failFlag = true
}
}
if failFlag {
status = constants.StatusFailed
}
return status, nil
}
func (r *ClickHouseReader) setColdStorage(ctx context.Context, tableName string, coldStorageVolume string) *model.ApiError {
// Set the storage policy for the required table. If it is already set, then setting it again
// will not a problem.
if len(coldStorageVolume) > 0 {
policyReq := fmt.Sprintf("ALTER TABLE %s ON CLUSTER %s MODIFY SETTING storage_policy='tiered'", tableName, r.cluster)
zap.L().Info("Executing Storage policy request: ", zap.String("request", policyReq))
if err := r.db.Exec(ctx, policyReq); err != nil {
zap.L().Error("error while setting storage policy", zap.Error(err))
return &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while setting storage policy. Err=%v", err)}
}
}
return nil
}
// GetDisks returns a list of disks {name, type} configured in clickhouse DB.
func (r *ClickHouseReader) GetDisks(ctx context.Context) (*[]model.DiskItem, *model.ApiError) {
diskItems := []model.DiskItem{}
query := "SELECT name,type FROM system.disks"
if err := r.db.Select(ctx, &diskItems, query); err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while getting disks. Err=%v", err)}
}
return &diskItems, nil
}
func getLocalTableNameArray(tableNames []string) []string {
var localTableNames []string
for _, name := range tableNames {
tableNameSplit := strings.Split(name, ".")
localTableNames = append(localTableNames, tableNameSplit[0]+"."+strings.Split(tableNameSplit[1], "distributed_")[1])
}
return localTableNames
}
// GetTTL returns current ttl, expected ttl and past setTTL status for metrics/traces.
func (r *ClickHouseReader) GetTTL(ctx context.Context, ttlParams *model.GetTTLParams) (*model.GetTTLResponseItem, *model.ApiError) {
parseTTL := func(queryResp string) (int, int) {
zap.L().Info("Parsing TTL from: ", zap.String("queryResp", queryResp))
deleteTTLExp := regexp.MustCompile(`toIntervalSecond\(([0-9]*)\)`)
moveTTLExp := regexp.MustCompile(`toIntervalSecond\(([0-9]*)\) TO VOLUME`)
var delTTL, moveTTL int = -1, -1
m := deleteTTLExp.FindStringSubmatch(queryResp)
if len(m) > 1 {
seconds_int, err := strconv.Atoi(m[1])
if err != nil {
return -1, -1
}
delTTL = seconds_int / 3600
}
m = moveTTLExp.FindStringSubmatch(queryResp)
if len(m) > 1 {
seconds_int, err := strconv.Atoi(m[1])
if err != nil {
return -1, -1
}
moveTTL = seconds_int / 3600
}
return delTTL, moveTTL
}
getMetricsTTL := func() (*model.DBResponseTTL, *model.ApiError) {
var dbResp []model.DBResponseTTL
query := fmt.Sprintf("SELECT engine_full FROM system.tables WHERE name='%v'", signozSampleLocalTableName)
err := r.db.Select(ctx, &dbResp, query)
if err != nil {
zap.L().Error("error while getting ttl", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while getting ttl. Err=%v", err)}
}
if len(dbResp) == 0 {
return nil, nil
} else {
return &dbResp[0], nil
}
}
getTracesTTL := func() (*model.DBResponseTTL, *model.ApiError) {
var dbResp []model.DBResponseTTL
query := fmt.Sprintf("SELECT engine_full FROM system.tables WHERE name='%v' AND database='%v'", r.traceLocalTableName, signozTraceDBName)
err := r.db.Select(ctx, &dbResp, query)
if err != nil {
zap.L().Error("error while getting ttl", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while getting ttl. Err=%v", err)}
}
if len(dbResp) == 0 {
return nil, nil
} else {
return &dbResp[0], nil
}
}
2022-08-04 14:28:10 +05:30
getLogsTTL := func() (*model.DBResponseTTL, *model.ApiError) {
var dbResp []model.DBResponseTTL
query := fmt.Sprintf("SELECT engine_full FROM system.tables WHERE name='%v' AND database='%v'", r.logsLocalTableName, r.logsDB)
2022-08-04 14:28:10 +05:30
err := r.db.Select(ctx, &dbResp, query)
if err != nil {
zap.L().Error("error while getting ttl", zap.Error(err))
2022-08-04 14:28:10 +05:30
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while getting ttl. Err=%v", err)}
}
if len(dbResp) == 0 {
return nil, nil
} else {
return &dbResp[0], nil
}
}
switch ttlParams.Type {
case constants.TraceTTL:
tableNameArray := []string{signozTraceDBName + "." + signozTraceTableName, signozTraceDBName + "." + signozDurationMVTable, signozTraceDBName + "." + signozSpansTable, signozTraceDBName + "." + signozErrorIndexTable, signozTraceDBName + "." + signozUsageExplorerTable, signozTraceDBName + "." + defaultDependencyGraphTable}
tableNameArray = getLocalTableNameArray(tableNameArray)
status, err := r.setTTLQueryStatus(ctx, tableNameArray)
if err != nil {
return nil, err
}
dbResp, err := getTracesTTL()
if err != nil {
return nil, err
}
ttlQuery, err := r.checkTTLStatusItem(ctx, tableNameArray[0])
if err != nil {
return nil, err
}
ttlQuery.TTL = ttlQuery.TTL / 3600 // convert to hours
if ttlQuery.ColdStorageTtl != -1 {
ttlQuery.ColdStorageTtl = ttlQuery.ColdStorageTtl / 3600 // convert to hours
}
delTTL, moveTTL := parseTTL(dbResp.EngineFull)
return &model.GetTTLResponseItem{TracesTime: delTTL, TracesMoveTime: moveTTL, ExpectedTracesTime: ttlQuery.TTL, ExpectedTracesMoveTime: ttlQuery.ColdStorageTtl, Status: status}, nil
case constants.MetricsTTL:
tableNameArray := []string{signozMetricDBName + "." + signozSampleTableName}
tableNameArray = getLocalTableNameArray(tableNameArray)
status, err := r.setTTLQueryStatus(ctx, tableNameArray)
if err != nil {
return nil, err
}
dbResp, err := getMetricsTTL()
if err != nil {
return nil, err
}
ttlQuery, err := r.checkTTLStatusItem(ctx, tableNameArray[0])
if err != nil {
return nil, err
}
ttlQuery.TTL = ttlQuery.TTL / 3600 // convert to hours
if ttlQuery.ColdStorageTtl != -1 {
ttlQuery.ColdStorageTtl = ttlQuery.ColdStorageTtl / 3600 // convert to hours
}
delTTL, moveTTL := parseTTL(dbResp.EngineFull)
return &model.GetTTLResponseItem{MetricsTime: delTTL, MetricsMoveTime: moveTTL, ExpectedMetricsTime: ttlQuery.TTL, ExpectedMetricsMoveTime: ttlQuery.ColdStorageTtl, Status: status}, nil
2022-08-04 14:28:10 +05:30
case constants.LogsTTL:
tableNameArray := []string{r.logsDB + "." + r.logsTable}
tableNameArray = getLocalTableNameArray(tableNameArray)
2022-08-04 14:28:10 +05:30
status, err := r.setTTLQueryStatus(ctx, tableNameArray)
if err != nil {
return nil, err
}
dbResp, err := getLogsTTL()
if err != nil {
return nil, err
}
ttlQuery, err := r.checkTTLStatusItem(ctx, tableNameArray[0])
if err != nil {
return nil, err
}
ttlQuery.TTL = ttlQuery.TTL / 3600 // convert to hours
if ttlQuery.ColdStorageTtl != -1 {
ttlQuery.ColdStorageTtl = ttlQuery.ColdStorageTtl / 3600 // convert to hours
}
delTTL, moveTTL := parseTTL(dbResp.EngineFull)
return &model.GetTTLResponseItem{LogsTime: delTTL, LogsMoveTime: moveTTL, ExpectedLogsTime: ttlQuery.TTL, ExpectedLogsMoveTime: ttlQuery.ColdStorageTtl, Status: status}, nil
default:
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error while getting ttl. ttl type should be metrics|traces, got %v",
ttlParams.Type)}
}
}
func (r *ClickHouseReader) ListErrors(ctx context.Context, queryParams *model.ListErrorsParams) (*[]model.Error, *model.ApiError) {
var getErrorResponses []model.Error
query := "SELECT any(exceptionMessage) as exceptionMessage, count() AS exceptionCount, min(timestamp) as firstSeen, max(timestamp) as lastSeen, groupID"
if len(queryParams.ServiceName) != 0 {
query = query + ", serviceName"
} else {
query = query + ", any(serviceName) as serviceName"
}
if len(queryParams.ExceptionType) != 0 {
query = query + ", exceptionType"
} else {
query = query + ", any(exceptionType) as exceptionType"
}
query += fmt.Sprintf(" FROM %s.%s WHERE timestamp >= @timestampL AND timestamp <= @timestampU", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("timestampL", strconv.FormatInt(queryParams.Start.UnixNano(), 10)), clickhouse.Named("timestampU", strconv.FormatInt(queryParams.End.UnixNano(), 10))}
if len(queryParams.ServiceName) != 0 {
query = query + " AND serviceName ilike @serviceName"
args = append(args, clickhouse.Named("serviceName", "%"+queryParams.ServiceName+"%"))
}
if len(queryParams.ExceptionType) != 0 {
query = query + " AND exceptionType ilike @exceptionType"
args = append(args, clickhouse.Named("exceptionType", "%"+queryParams.ExceptionType+"%"))
}
// create TagQuery from TagQueryParams
tags := createTagQueryFromTagQueryParams(queryParams.Tags)
subQuery, argsSubQuery, errStatus := buildQueryWithTagParams(ctx, tags)
query += subQuery
args = append(args, argsSubQuery...)
if errStatus != nil {
zap.L().Error("Error in processing tags", zap.Error(errStatus))
return nil, errStatus
}
query = query + " GROUP BY groupID"
if len(queryParams.ServiceName) != 0 {
query = query + ", serviceName"
}
if len(queryParams.ExceptionType) != 0 {
query = query + ", exceptionType"
}
if len(queryParams.OrderParam) != 0 {
if queryParams.Order == constants.Descending {
query = query + " ORDER BY " + queryParams.OrderParam + " DESC"
} else if queryParams.Order == constants.Ascending {
query = query + " ORDER BY " + queryParams.OrderParam + " ASC"
}
}
if queryParams.Limit > 0 {
query = query + " LIMIT @limit"
args = append(args, clickhouse.Named("limit", queryParams.Limit))
}
if queryParams.Offset > 0 {
query = query + " OFFSET @offset"
args = append(args, clickhouse.Named("offset", queryParams.Offset))
}
err := r.db.Select(ctx, &getErrorResponses, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
return &getErrorResponses, nil
}
func (r *ClickHouseReader) CountErrors(ctx context.Context, queryParams *model.CountErrorsParams) (uint64, *model.ApiError) {
var errorCount uint64
query := fmt.Sprintf("SELECT count(distinct(groupID)) FROM %s.%s WHERE timestamp >= @timestampL AND timestamp <= @timestampU", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("timestampL", strconv.FormatInt(queryParams.Start.UnixNano(), 10)), clickhouse.Named("timestampU", strconv.FormatInt(queryParams.End.UnixNano(), 10))}
if len(queryParams.ServiceName) != 0 {
query = query + " AND serviceName ilike @serviceName"
args = append(args, clickhouse.Named("serviceName", "%"+queryParams.ServiceName+"%"))
}
if len(queryParams.ExceptionType) != 0 {
query = query + " AND exceptionType ilike @exceptionType"
args = append(args, clickhouse.Named("exceptionType", "%"+queryParams.ExceptionType+"%"))
}
// create TagQuery from TagQueryParams
tags := createTagQueryFromTagQueryParams(queryParams.Tags)
subQuery, argsSubQuery, errStatus := buildQueryWithTagParams(ctx, tags)
query += subQuery
args = append(args, argsSubQuery...)
if errStatus != nil {
zap.L().Error("Error in processing tags", zap.Error(errStatus))
return 0, errStatus
}
err := r.db.QueryRow(ctx, query, args...).Scan(&errorCount)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return 0, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
return errorCount, nil
}
func (r *ClickHouseReader) GetErrorFromErrorID(ctx context.Context, queryParams *model.GetErrorParams) (*model.ErrorWithSpan, *model.ApiError) {
if queryParams.ErrorID == "" {
zap.L().Error("errorId missing from params")
return nil, &model.ApiError{Typ: model.ErrorBadData, Err: fmt.Errorf("ErrorID missing from params")}
}
var getErrorWithSpanReponse []model.ErrorWithSpan
query := fmt.Sprintf("SELECT errorID, exceptionType, exceptionStacktrace, exceptionEscaped, exceptionMessage, timestamp, spanID, traceID, serviceName, groupID FROM %s.%s WHERE timestamp = @timestamp AND groupID = @groupID AND errorID = @errorID LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getErrorWithSpanReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getErrorWithSpanReponse) > 0 {
return &getErrorWithSpanReponse[0], nil
} else {
return nil, &model.ApiError{Typ: model.ErrorNotFound, Err: fmt.Errorf("Error/Exception not found")}
}
}
func (r *ClickHouseReader) GetErrorFromGroupID(ctx context.Context, queryParams *model.GetErrorParams) (*model.ErrorWithSpan, *model.ApiError) {
var getErrorWithSpanReponse []model.ErrorWithSpan
query := fmt.Sprintf("SELECT errorID, exceptionType, exceptionStacktrace, exceptionEscaped, exceptionMessage, timestamp, spanID, traceID, serviceName, groupID FROM %s.%s WHERE timestamp = @timestamp AND groupID = @groupID LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getErrorWithSpanReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getErrorWithSpanReponse) > 0 {
return &getErrorWithSpanReponse[0], nil
} else {
return nil, &model.ApiError{Typ: model.ErrorNotFound, Err: fmt.Errorf("Error/Exception not found")}
}
}
func (r *ClickHouseReader) GetNextPrevErrorIDs(ctx context.Context, queryParams *model.GetErrorParams) (*model.NextPrevErrorIDs, *model.ApiError) {
if queryParams.ErrorID == "" {
zap.L().Error("errorId missing from params")
return nil, &model.ApiError{Typ: model.ErrorBadData, Err: fmt.Errorf("ErrorID missing from params")}
}
var err *model.ApiError
getNextPrevErrorIDsResponse := model.NextPrevErrorIDs{
GroupID: queryParams.GroupID,
}
getNextPrevErrorIDsResponse.NextErrorID, getNextPrevErrorIDsResponse.NextTimestamp, err = r.getNextErrorID(ctx, queryParams)
if err != nil {
zap.L().Error("Unable to get next error ID due to err: ", zap.Error(err))
return nil, err
}
getNextPrevErrorIDsResponse.PrevErrorID, getNextPrevErrorIDsResponse.PrevTimestamp, err = r.getPrevErrorID(ctx, queryParams)
if err != nil {
zap.L().Error("Unable to get prev error ID due to err: ", zap.Error(err))
return nil, err
}
return &getNextPrevErrorIDsResponse, nil
}
func (r *ClickHouseReader) getNextErrorID(ctx context.Context, queryParams *model.GetErrorParams) (string, time.Time, *model.ApiError) {
var getNextErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as nextErrorID, timestamp as nextTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp >= @timestamp AND errorID != @errorID ORDER BY timestamp ASC LIMIT 2", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getNextErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getNextErrorIDReponse) == 0 {
zap.L().Info("NextErrorID not found")
return "", time.Time{}, nil
} else if len(getNextErrorIDReponse) == 1 {
zap.L().Info("NextErrorID found")
return getNextErrorIDReponse[0].NextErrorID, getNextErrorIDReponse[0].NextTimestamp, nil
} else {
if getNextErrorIDReponse[0].Timestamp.UnixNano() == getNextErrorIDReponse[1].Timestamp.UnixNano() {
var getNextErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as nextErrorID, timestamp as nextTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp = @timestamp AND errorID > @errorID ORDER BY errorID ASC LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getNextErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getNextErrorIDReponse) == 0 {
var getNextErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as nextErrorID, timestamp as nextTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp > @timestamp ORDER BY timestamp ASC LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getNextErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getNextErrorIDReponse) == 0 {
zap.L().Info("NextErrorID not found")
return "", time.Time{}, nil
} else {
zap.L().Info("NextErrorID found")
return getNextErrorIDReponse[0].NextErrorID, getNextErrorIDReponse[0].NextTimestamp, nil
}
} else {
zap.L().Info("NextErrorID found")
return getNextErrorIDReponse[0].NextErrorID, getNextErrorIDReponse[0].NextTimestamp, nil
}
} else {
zap.L().Info("NextErrorID found")
return getNextErrorIDReponse[0].NextErrorID, getNextErrorIDReponse[0].NextTimestamp, nil
}
}
}
func (r *ClickHouseReader) getPrevErrorID(ctx context.Context, queryParams *model.GetErrorParams) (string, time.Time, *model.ApiError) {
var getPrevErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as prevErrorID, timestamp as prevTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp <= @timestamp AND errorID != @errorID ORDER BY timestamp DESC LIMIT 2", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getPrevErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getPrevErrorIDReponse) == 0 {
zap.L().Info("PrevErrorID not found")
return "", time.Time{}, nil
} else if len(getPrevErrorIDReponse) == 1 {
zap.L().Info("PrevErrorID found")
return getPrevErrorIDReponse[0].PrevErrorID, getPrevErrorIDReponse[0].PrevTimestamp, nil
} else {
if getPrevErrorIDReponse[0].Timestamp.UnixNano() == getPrevErrorIDReponse[1].Timestamp.UnixNano() {
var getPrevErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as prevErrorID, timestamp as prevTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp = @timestamp AND errorID < @errorID ORDER BY errorID DESC LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getPrevErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getPrevErrorIDReponse) == 0 {
var getPrevErrorIDReponse []model.NextPrevErrorIDsDBResponse
query := fmt.Sprintf("SELECT errorID as prevErrorID, timestamp as prevTimestamp FROM %s.%s WHERE groupID = @groupID AND timestamp < @timestamp ORDER BY timestamp DESC LIMIT 1", r.TraceDB, r.errorTable)
args := []interface{}{clickhouse.Named("errorID", queryParams.ErrorID), clickhouse.Named("groupID", queryParams.GroupID), clickhouse.Named("timestamp", strconv.FormatInt(queryParams.Timestamp.UnixNano(), 10))}
err := r.db.Select(ctx, &getPrevErrorIDReponse, query, args...)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return "", time.Time{}, &model.ApiError{Typ: model.ErrorExec, Err: fmt.Errorf("error in processing sql query")}
}
if len(getPrevErrorIDReponse) == 0 {
zap.L().Info("PrevErrorID not found")
return "", time.Time{}, nil
} else {
zap.L().Info("PrevErrorID found")
return getPrevErrorIDReponse[0].PrevErrorID, getPrevErrorIDReponse[0].PrevTimestamp, nil
}
} else {
zap.L().Info("PrevErrorID found")
return getPrevErrorIDReponse[0].PrevErrorID, getPrevErrorIDReponse[0].PrevTimestamp, nil
}
} else {
zap.L().Info("PrevErrorID found")
return getPrevErrorIDReponse[0].PrevErrorID, getPrevErrorIDReponse[0].PrevTimestamp, nil
}
}
}
func (r *ClickHouseReader) GetTotalSpans(ctx context.Context) (uint64, error) {
var totalSpans uint64
queryStr := fmt.Sprintf("SELECT count() from %s.%s;", signozTraceDBName, r.traceTableName)
r.db.QueryRow(ctx, queryStr).Scan(&totalSpans)
return totalSpans, nil
}
func (r *ClickHouseReader) GetSpansInLastHeartBeatInterval(ctx context.Context, interval time.Duration) (uint64, error) {
var spansInLastHeartBeatInterval uint64
queryStr := fmt.Sprintf("SELECT count() from %s.%s where timestamp > toUnixTimestamp(now()-toIntervalMinute(%d));", signozTraceDBName, signozSpansTable, int(interval.Minutes()))
if r.useTraceNewSchema {
queryStr = fmt.Sprintf("SELECT count() from %s.%s where ts_bucket_start >= toUInt64(toUnixTimestamp(now() - toIntervalMinute(%d))) - 1800 and timestamp > toUnixTimestamp(now()-toIntervalMinute(%d));", signozTraceDBName, r.traceTableName, int(interval.Minutes()), int(interval.Minutes()))
}
r.db.QueryRow(ctx, queryStr).Scan(&spansInLastHeartBeatInterval)
return spansInLastHeartBeatInterval, nil
}
func (r *ClickHouseReader) GetTotalLogs(ctx context.Context) (uint64, error) {
var totalLogs uint64
queryStr := fmt.Sprintf("SELECT count() from %s.%s;", r.logsDB, r.logsTableName)
r.db.QueryRow(ctx, queryStr).Scan(&totalLogs)
return totalLogs, nil
}
func (r *ClickHouseReader) FetchTemporality(ctx context.Context, metricNames []string) (map[string]map[v3.Temporality]bool, error) {
metricNameToTemporality := make(map[string]map[v3.Temporality]bool)
query := fmt.Sprintf(`SELECT DISTINCT metric_name, temporality FROM %s.%s WHERE metric_name IN $1`, signozMetricDBName, signozTSTableNameV41Day)
rows, err := r.db.Query(ctx, query, metricNames)
if err != nil {
return nil, err
}
defer rows.Close()
for rows.Next() {
var metricName, temporality string
err := rows.Scan(&metricName, &temporality)
if err != nil {
return nil, err
}
if _, ok := metricNameToTemporality[metricName]; !ok {
metricNameToTemporality[metricName] = make(map[v3.Temporality]bool)
}
metricNameToTemporality[metricName][v3.Temporality(temporality)] = true
}
return metricNameToTemporality, nil
}
func (r *ClickHouseReader) GetTimeSeriesInfo(ctx context.Context) (map[string]interface{}, error) {
queryStr := fmt.Sprintf("SELECT countDistinct(fingerprint) as count from %s.%s where metric_name not like 'signoz_%%' group by metric_name order by count desc;", signozMetricDBName, signozTSTableNameV41Day)
rows, _ := r.db.Query(ctx, queryStr)
var totalTS uint64
totalTS = 0
var maxTS uint64
maxTS = 0
count := 0
for rows.Next() {
var value uint64
rows.Scan(&value)
totalTS += value
if count == 0 {
maxTS = value
}
count += 1
}
timeSeriesData := map[string]interface{}{}
timeSeriesData["totalTS"] = totalTS
timeSeriesData["maxTS"] = maxTS
return timeSeriesData, nil
}
func (r *ClickHouseReader) GetSamplesInfoInLastHeartBeatInterval(ctx context.Context, interval time.Duration) (uint64, error) {
var totalSamples uint64
queryStr := fmt.Sprintf("select count() from %s.%s where metric_name not like 'signoz_%%' and unix_milli > toUnixTimestamp(now()-toIntervalMinute(%d))*1000;", signozMetricDBName, signozSampleTableName, int(interval.Minutes()))
r.db.QueryRow(ctx, queryStr).Scan(&totalSamples)
return totalSamples, nil
}
func (r *ClickHouseReader) GetTotalSamples(ctx context.Context) (uint64, error) {
var totalSamples uint64
queryStr := fmt.Sprintf("select count() from %s.%s where metric_name not like 'signoz_%%';", signozMetricDBName, signozSampleTableName)
r.db.QueryRow(ctx, queryStr).Scan(&totalSamples)
return totalSamples, nil
}
func (r *ClickHouseReader) GetDistributedInfoInLastHeartBeatInterval(ctx context.Context) (map[string]interface{}, error) {
clusterInfo := []model.ClusterInfo{}
queryStr := `SELECT shard_num, shard_weight, replica_num, errors_count, slowdowns_count, estimated_recovery_time FROM system.clusters where cluster='cluster';`
r.db.Select(ctx, &clusterInfo, queryStr)
if len(clusterInfo) == 1 {
return clusterInfo[0].GetMapFromStruct(), nil
}
return nil, nil
}
func (r *ClickHouseReader) GetLogsInfoInLastHeartBeatInterval(ctx context.Context, interval time.Duration) (uint64, error) {
2022-08-11 14:27:19 +05:30
var totalLogLines uint64
2025-01-14 01:03:12 +05:30
queryStr := fmt.Sprintf("select count() from %s.%s where timestamp > toUnixTimestamp(now()-toIntervalMinute(%d))*1000000000;", r.logsDB, r.logsTableV2, int(interval.Minutes()))
2022-08-11 14:27:19 +05:30
err := r.db.QueryRow(ctx, queryStr).Scan(&totalLogLines)
2022-08-11 14:27:19 +05:30
return totalLogLines, err
2022-08-11 14:27:19 +05:30
}
2022-07-12 16:38:26 +05:30
func (r *ClickHouseReader) GetTagsInfoInLastHeartBeatInterval(ctx context.Context, interval time.Duration) (*model.TagsInfo, error) {
queryStr := fmt.Sprintf(`select serviceName, stringTagMap['deployment.environment'] as env,
stringTagMap['telemetry.sdk.language'] as language from %s.%s
where timestamp > toUnixTimestamp(now()-toIntervalMinute(%d))
group by serviceName, env, language;`, r.TraceDB, r.traceTableName, int(interval.Minutes()))
if r.useTraceNewSchema {
queryStr = fmt.Sprintf(`select serviceName, resources_string['deployment.environment'] as env,
resources_string['telemetry.sdk.language'] as language from %s.%s
where timestamp > toUnixTimestamp(now()-toIntervalMinute(%d))
group by serviceName, env, language;`, r.TraceDB, r.traceTableName, int(interval.Minutes()))
}
2022-10-11 00:43:54 +05:30
tagTelemetryDataList := []model.TagTelemetryData{}
err := r.db.Select(ctx, &tagTelemetryDataList, queryStr)
if err != nil {
zap.L().Error("Error in processing sql query: ", zap.Error(err))
2022-10-11 00:43:54 +05:30
return nil, err
}
tagsInfo := model.TagsInfo{
Languages: make(map[string]interface{}),
Services: make(map[string]interface{}),
2022-10-11 00:43:54 +05:30
}
for _, tagTelemetryData := range tagTelemetryDataList {
if len(tagTelemetryData.ServiceName) != 0 && strings.Contains(tagTelemetryData.ServiceName, "prod") {
tagsInfo.Env = tagTelemetryData.ServiceName
}
if len(tagTelemetryData.Env) != 0 && strings.Contains(tagTelemetryData.Env, "prod") {
tagsInfo.Env = tagTelemetryData.Env
}
if len(tagTelemetryData.Language) != 0 {
tagsInfo.Languages[tagTelemetryData.Language] = struct{}{}
}
if len(tagTelemetryData.ServiceName) != 0 {
tagsInfo.Services[tagTelemetryData.ServiceName] = struct{}{}
}
2022-10-11 00:43:54 +05:30
}
return &tagsInfo, nil
}
// remove this after sometime
func removeUnderscoreDuplicateFields(fields []model.Field) []model.Field {
lookup := map[string]model.Field{}
for _, v := range fields {
lookup[v.Name+v.DataType] = v
}
for k := range lookup {
if strings.Contains(k, ".") {
delete(lookup, strings.ReplaceAll(k, ".", "_"))
}
}
updatedFields := []model.Field{}
for _, v := range lookup {
updatedFields = append(updatedFields, v)
}
return updatedFields
}
2022-07-12 16:38:26 +05:30
func (r *ClickHouseReader) GetLogFields(ctx context.Context) (*model.GetFieldsResponse, *model.ApiError) {
// response will contain top level fields from the otel log model
response := model.GetFieldsResponse{
Selected: constants.StaticSelectedLogFields,
Interesting: []model.Field{},
2022-07-12 16:38:26 +05:30
}
// get attribute keys
attributes := []model.Field{}
2022-07-12 16:38:26 +05:30
query := fmt.Sprintf("SELECT DISTINCT name, datatype from %s.%s group by name, datatype", r.logsDB, r.logsAttributeKeys)
err := r.db.Select(ctx, &attributes, query)
2022-07-12 16:38:26 +05:30
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
// get resource keys
resources := []model.Field{}
2022-07-12 16:38:26 +05:30
query = fmt.Sprintf("SELECT DISTINCT name, datatype from %s.%s group by name, datatype", r.logsDB, r.logsResourceKeys)
err = r.db.Select(ctx, &resources, query)
2022-07-12 16:38:26 +05:30
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
//remove this code after sometime
attributes = removeUnderscoreDuplicateFields(attributes)
resources = removeUnderscoreDuplicateFields(resources)
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.logsDB, r.logsLocalTableName)
2022-07-12 16:38:26 +05:30
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
r.extractSelectedAndInterestingFields(statements[0].Statement, constants.Attributes, &attributes, &response)
r.extractSelectedAndInterestingFields(statements[0].Statement, constants.Resources, &resources, &response)
2022-07-12 16:38:26 +05:30
return &response, nil
}
func (r *ClickHouseReader) extractSelectedAndInterestingFields(tableStatement string, overrideFieldType string, fields *[]model.Field, response *model.GetFieldsResponse) {
2022-07-12 16:38:26 +05:30
for _, field := range *fields {
if overrideFieldType != "" {
field.Type = overrideFieldType
}
// all static fields are assumed to be selected as we don't allow changing them
if isColumn(r.useLogsNewSchema, tableStatement, field.Type, field.Name, field.DataType) {
2022-07-12 16:38:26 +05:30
response.Selected = append(response.Selected, field)
} else {
response.Interesting = append(response.Interesting, field)
}
}
}
func (r *ClickHouseReader) UpdateLogFieldV2(ctx context.Context, field *model.UpdateField) *model.ApiError {
if !field.Selected {
return model.ForbiddenError(errors.New("removing a selected field is not allowed, please reach out to support."))
}
colname := utils.GetClickhouseColumnNameV2(field.Type, field.DataType, field.Name)
field.DataType = strings.ToLower(field.DataType)
dataType := constants.MaterializedDataTypeMap[field.DataType]
chDataType := constants.ChDataTypeMap[field.DataType]
attrColName := fmt.Sprintf("%s_%s", field.Type, dataType)
for _, table := range []string{r.logsLocalTableV2, r.logsTableV2} {
q := "ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS `%s` %s DEFAULT %s['%s'] CODEC(ZSTD(1))"
query := fmt.Sprintf(q,
r.logsDB, table,
r.cluster,
colname, chDataType,
attrColName,
field.Name,
)
err := r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
query = fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS `%s_exists` bool DEFAULT if(mapContains(%s, '%s') != 0, true, false) CODEC(ZSTD(1))",
r.logsDB, table,
r.cluster,
colname,
attrColName,
field.Name,
)
err = r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
}
// create the index
if strings.ToLower(field.DataType) == "bool" {
// there is no point in creating index for bool attributes as the cardinality is just 2
return nil
}
if field.IndexType == "" {
field.IndexType = constants.DefaultLogSkipIndexType
}
if field.IndexGranularity == 0 {
field.IndexGranularity = constants.DefaultLogSkipIndexGranularity
}
query := fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD INDEX IF NOT EXISTS `%s_idx` (`%s`) TYPE %s GRANULARITY %d",
r.logsDB, r.logsLocalTableV2,
r.cluster,
colname,
colname,
field.IndexType,
field.IndexGranularity,
)
err := r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
return nil
}
2022-07-12 16:38:26 +05:30
func (r *ClickHouseReader) UpdateLogField(ctx context.Context, field *model.UpdateField) *model.ApiError {
// don't allow updating static fields
if field.Type == constants.Static {
err := errors.New("cannot update static fields")
return &model.ApiError{Err: err, Typ: model.ErrorBadData}
}
if r.useLogsNewSchema {
return r.UpdateLogFieldV2(ctx, field)
}
// if a field is selected it means that the field needs to be indexed
2022-07-12 16:38:26 +05:30
if field.Selected {
colname := utils.GetClickhouseColumnName(field.Type, field.DataType, field.Name)
keyColName := fmt.Sprintf("%s_%s_key", field.Type, strings.ToLower(field.DataType))
valueColName := fmt.Sprintf("%s_%s_value", field.Type, strings.ToLower(field.DataType))
// create materialized column
for _, table := range []string{r.logsLocalTable, r.logsTable} {
q := "ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS %s %s DEFAULT %s[indexOf(%s, '%s')] CODEC(ZSTD(1))"
query := fmt.Sprintf(q,
r.logsDB, table,
r.cluster,
colname, field.DataType,
valueColName,
keyColName,
field.Name,
)
err := r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
query = fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS %s_exists` bool DEFAULT if(indexOf(%s, '%s') != 0, true, false) CODEC(ZSTD(1))",
r.logsDB, table,
r.cluster,
strings.TrimSuffix(colname, "`"),
keyColName,
field.Name,
)
err = r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
}
// create the index
if strings.ToLower(field.DataType) == "bool" {
// there is no point in creating index for bool attributes as the cardinality is just 2
return nil
2022-07-12 16:38:26 +05:30
}
2022-07-22 16:49:40 +05:30
if field.IndexType == "" {
field.IndexType = constants.DefaultLogSkipIndexType
2022-07-12 16:38:26 +05:30
}
2022-07-22 16:49:40 +05:30
if field.IndexGranularity == 0 {
field.IndexGranularity = constants.DefaultLogSkipIndexGranularity
2022-07-12 16:38:26 +05:30
}
query := fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD INDEX IF NOT EXISTS %s_idx` (%s) TYPE %s GRANULARITY %d",
r.logsDB, r.logsLocalTable,
r.cluster,
strings.TrimSuffix(colname, "`"),
colname,
field.IndexType,
field.IndexGranularity,
)
err := r.db.Exec(ctx, query)
2022-07-12 16:38:26 +05:30
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
2022-07-12 16:38:26 +05:30
} else {
// We are not allowing to delete a materialized column
// For more details please check https://github.com/SigNoz/signoz/issues/4566
return model.ForbiddenError(errors.New("Removing a selected field is not allowed, please reach out to support."))
// Delete the index first
// query := fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s DROP INDEX IF EXISTS %s_idx`", r.logsDB, r.logsLocalTable, r.cluster, strings.TrimSuffix(colname, "`"))
// err := r.db.Exec(ctx, query)
// if err != nil {
// return &model.ApiError{Err: err, Typ: model.ErrorInternal}
// }
// for _, table := range []string{r.logsTable, r.logsLocalTable} {
// // drop materialized column from logs table
// query := "ALTER TABLE %s.%s ON CLUSTER %s DROP COLUMN IF EXISTS %s "
// err := r.db.Exec(ctx, fmt.Sprintf(query,
// r.logsDB, table,
// r.cluster,
// colname,
// ),
// )
// if err != nil {
// return &model.ApiError{Err: err, Typ: model.ErrorInternal}
// }
// // drop exists column on logs table
// query = "ALTER TABLE %s.%s ON CLUSTER %s DROP COLUMN IF EXISTS %s_exists` "
// err = r.db.Exec(ctx, fmt.Sprintf(query,
// r.logsDB, table,
// r.cluster,
// strings.TrimSuffix(colname, "`"),
// ),
// )
// if err != nil {
// return &model.ApiError{Err: err, Typ: model.ErrorInternal}
// }
// }
2022-07-12 16:38:26 +05:30
}
return nil
}
func (r *ClickHouseReader) GetTraceFields(ctx context.Context) (*model.GetFieldsResponse, *model.ApiError) {
// response will contain top level fields from the otel trace model
response := model.GetFieldsResponse{
Selected: []model.Field{},
Interesting: []model.Field{},
}
// get the top level selected fields
for _, field := range constants.NewStaticFieldsTraces {
if (v3.AttributeKey{} == field) {
continue
}
response.Selected = append(response.Selected, model.Field{
Name: field.Key,
DataType: field.DataType.String(),
Type: constants.Static,
})
}
// get attribute keys
attributes := []model.Field{}
query := fmt.Sprintf("SELECT tagKey, tagType, dataType from %s.%s group by tagKey, tagType, dataType", r.TraceDB, r.spanAttributesKeysTable)
rows, err := r.db.Query(ctx, query)
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
defer rows.Close()
var tagKey string
var dataType string
var tagType string
for rows.Next() {
if err := rows.Scan(&tagKey, &tagType, &dataType); err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
attributes = append(attributes, model.Field{
Name: tagKey,
DataType: dataType,
Type: tagType,
})
}
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.TraceDB, r.traceLocalTableName)
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
r.extractSelectedAndInterestingFields(statements[0].Statement, "", &attributes, &response)
return &response, nil
}
func (r *ClickHouseReader) UpdateTraceField(ctx context.Context, field *model.UpdateField) *model.ApiError {
if !field.Selected {
return model.ForbiddenError(errors.New("removing a selected field is not allowed, please reach out to support."))
}
// name of the materialized column
colname := utils.GetClickhouseColumnNameV2(field.Type, field.DataType, field.Name)
field.DataType = strings.ToLower(field.DataType)
// dataType and chDataType of the materialized column
chDataType := constants.ChDataTypeMap[field.DataType]
dataType := constants.MaterializedDataTypeMap[field.DataType]
// typeName: tag => attributes, resource => resources
typeName := field.Type
if field.Type == string(v3.AttributeKeyTypeTag) {
typeName = constants.Attributes
} else if field.Type == string(v3.AttributeKeyTypeResource) {
typeName = constants.Resources
}
attrColName := fmt.Sprintf("%s_%s", typeName, dataType)
for _, table := range []string{r.traceLocalTableName, r.traceTableName} {
q := "ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS `%s` %s DEFAULT %s['%s'] CODEC(ZSTD(1))"
query := fmt.Sprintf(q,
r.TraceDB, table,
r.cluster,
colname, chDataType,
attrColName,
field.Name,
)
err := r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
query = fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD COLUMN IF NOT EXISTS `%s_exists` bool DEFAULT if(mapContains(%s, '%s') != 0, true, false) CODEC(ZSTD(1))",
r.TraceDB, table,
r.cluster,
colname,
attrColName,
field.Name,
)
err = r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
}
// create the index
if strings.ToLower(field.DataType) == "bool" {
// there is no point in creating index for bool attributes as the cardinality is just 2
return nil
}
if field.IndexType == "" {
field.IndexType = constants.DefaultLogSkipIndexType
}
if field.IndexGranularity == 0 {
field.IndexGranularity = constants.DefaultLogSkipIndexGranularity
}
query := fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD INDEX IF NOT EXISTS `%s_idx` (`%s`) TYPE %s GRANULARITY %d",
r.TraceDB, r.traceLocalTableName,
r.cluster,
colname,
colname,
field.IndexType,
field.IndexGranularity,
)
err := r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
// add a default minmax index for numbers
if dataType == "number" {
query = fmt.Sprintf("ALTER TABLE %s.%s ON CLUSTER %s ADD INDEX IF NOT EXISTS `%s_minmax_idx` (`%s`) TYPE minmax GRANULARITY 1",
r.TraceDB, r.traceLocalTableName,
r.cluster,
colname,
colname,
)
err = r.db.Exec(ctx, query)
if err != nil {
return &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
}
return nil
}
func (r *ClickHouseReader) GetLogs(ctx context.Context, params *model.LogsFilterParams) (*[]model.SignozLog, *model.ApiError) {
response := []model.SignozLog{}
fields, apiErr := r.GetLogFields(ctx)
if apiErr != nil {
return nil, apiErr
}
2022-08-10 14:27:46 +05:30
isPaginatePrev := logs.CheckIfPrevousPaginateAndModifyOrder(params)
filterSql, lenFilters, err := logs.GenerateSQLWhere(fields, params)
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorBadData}
}
data := map[string]interface{}{
"lenFilters": lenFilters,
}
if lenFilters != 0 {
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_LOGS_FILTERS, data, claims.Email, true, false)
}
}
query := fmt.Sprintf("%s from %s.%s", constants.LogsSQLSelect, r.logsDB, r.logsTable)
if filterSql != "" {
2022-07-27 10:46:33 +05:30
query = fmt.Sprintf("%s where %s", query, filterSql)
}
query = fmt.Sprintf("%s order by %s %s limit %d", query, params.OrderBy, params.Order, params.Limit)
err = r.db.Select(ctx, &response, query)
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
2022-08-10 14:27:46 +05:30
if isPaginatePrev {
// rever the results from db
for i, j := 0, len(response)-1; i < j; i, j = i+1, j-1 {
response[i], response[j] = response[j], response[i]
}
}
return &response, nil
}
2022-07-18 16:37:46 +05:30
2022-07-18 18:55:52 +05:30
func (r *ClickHouseReader) TailLogs(ctx context.Context, client *model.LogsTailClient) {
2022-07-19 16:34:33 +05:30
2022-07-18 18:55:52 +05:30
fields, apiErr := r.GetLogFields(ctx)
if apiErr != nil {
client.Error <- apiErr.Err
return
}
filterSql, lenFilters, err := logs.GenerateSQLWhere(fields, &model.LogsFilterParams{
2022-07-18 18:55:52 +05:30
Query: client.Filter.Query,
})
data := map[string]interface{}{
"lenFilters": lenFilters,
}
if lenFilters != 0 {
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_LOGS_FILTERS, data, claims.Email, true, false)
}
}
2022-07-18 18:55:52 +05:30
if err != nil {
client.Error <- err
return
}
query := fmt.Sprintf("%s from %s.%s", constants.LogsSQLSelect, r.logsDB, r.logsTable)
2022-07-18 18:55:52 +05:30
2022-07-22 15:49:50 +05:30
tsStart := uint64(time.Now().UnixNano())
2022-07-22 16:49:40 +05:30
if client.Filter.TimestampStart != 0 {
tsStart = client.Filter.TimestampStart
2022-07-18 18:55:52 +05:30
}
var idStart string
2022-08-10 14:27:46 +05:30
if client.Filter.IdGt != "" {
idStart = client.Filter.IdGt
2022-07-18 18:55:52 +05:30
}
ticker := time.NewTicker(time.Duration(r.liveTailRefreshSeconds) * time.Second)
defer ticker.Stop()
2022-07-18 18:55:52 +05:30
for {
2022-07-18 16:37:46 +05:30
select {
case <-ctx.Done():
2022-07-18 18:55:52 +05:30
done := true
client.Done <- &done
zap.L().Debug("closing go routine : " + client.Name)
2022-07-18 18:55:52 +05:30
return
case <-ticker.C:
// get the new 100 logs as anything more older won't make sense
tmpQuery := fmt.Sprintf("%s where timestamp >='%d'", query, tsStart)
if filterSql != "" {
2022-07-27 10:46:33 +05:30
tmpQuery = fmt.Sprintf("%s and %s", tmpQuery, filterSql)
2022-07-18 18:55:52 +05:30
}
if idStart != "" {
2022-07-27 10:46:33 +05:30
tmpQuery = fmt.Sprintf("%s and id > '%s'", tmpQuery, idStart)
2022-07-18 18:55:52 +05:30
}
tmpQuery = fmt.Sprintf("%s order by timestamp desc, id desc limit 100", tmpQuery)
response := []model.SignozLog{}
err := r.db.Select(ctx, &response, tmpQuery)
2022-07-18 18:55:52 +05:30
if err != nil {
zap.L().Error("Error while getting logs", zap.Error(err))
2022-07-18 18:55:52 +05:30
client.Error <- err
return
}
2022-07-27 10:46:33 +05:30
for i := len(response) - 1; i >= 0; i-- {
2022-07-18 18:55:52 +05:30
select {
case <-ctx.Done():
done := true
client.Done <- &done
zap.L().Debug("closing go routine while sending logs : " + client.Name)
2022-07-18 18:55:52 +05:30
return
default:
client.Logs <- &response[i]
if i == 0 {
tsStart = response[i].Timestamp
idStart = response[i].ID
2022-07-18 18:55:52 +05:30
}
}
}
2022-07-18 16:37:46 +05:30
}
}
}
2022-07-20 12:11:03 +05:30
func (r *ClickHouseReader) AggregateLogs(ctx context.Context, params *model.LogsAggregateParams) (*model.GetLogsAggregatesResponse, *model.ApiError) {
logAggregatesDBResponseItems := []model.LogsAggregatesDBResponseItem{}
2022-07-20 12:11:03 +05:30
function := "toFloat64(count()) as value"
2022-07-22 16:49:40 +05:30
if params.Function != "" {
function = fmt.Sprintf("toFloat64(%s) as value", params.Function)
2022-07-20 12:11:03 +05:30
}
fields, apiErr := r.GetLogFields(ctx)
if apiErr != nil {
return nil, apiErr
}
filterSql, lenFilters, err := logs.GenerateSQLWhere(fields, &model.LogsFilterParams{
2022-07-20 12:11:03 +05:30
Query: params.Query,
})
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorBadData}
}
data := map[string]interface{}{
"lenFilters": lenFilters,
}
if lenFilters != 0 {
claims, ok := authtypes.ClaimsFromContext(ctx)
if ok {
telemetry.GetInstance().SendEvent(telemetry.TELEMETRY_EVENT_LOGS_FILTERS, data, claims.Email, true, false)
}
}
2022-07-20 12:11:03 +05:30
query := ""
2022-07-22 16:49:40 +05:30
if params.GroupBy != "" {
query = fmt.Sprintf("SELECT toInt64(toUnixTimestamp(toStartOfInterval(toDateTime(timestamp/1000000000), INTERVAL %d minute))*1000000000) as ts_start_interval, toString(%s) as groupBy, "+
2022-07-20 12:11:03 +05:30
"%s "+
"FROM %s.%s WHERE (timestamp >= '%d' AND timestamp <= '%d' )",
2022-07-22 16:49:40 +05:30
params.StepSeconds/60, params.GroupBy, function, r.logsDB, r.logsTable, params.TimestampStart, params.TimestampEnd)
2022-07-20 12:11:03 +05:30
} else {
query = fmt.Sprintf("SELECT toInt64(toUnixTimestamp(toStartOfInterval(toDateTime(timestamp/1000000000), INTERVAL %d minute))*1000000000) as ts_start_interval, "+
2022-07-20 12:11:03 +05:30
"%s "+
"FROM %s.%s WHERE (timestamp >= '%d' AND timestamp <= '%d' )",
2022-07-22 16:49:40 +05:30
params.StepSeconds/60, function, r.logsDB, r.logsTable, params.TimestampStart, params.TimestampEnd)
2022-07-20 12:11:03 +05:30
}
if filterSql != "" {
query = fmt.Sprintf("%s AND ( %s ) ", query, filterSql)
2022-07-20 12:11:03 +05:30
}
2022-07-22 16:49:40 +05:30
if params.GroupBy != "" {
query = fmt.Sprintf("%s GROUP BY ts_start_interval, toString(%s) as groupBy ORDER BY ts_start_interval", query, params.GroupBy)
2022-07-20 12:11:03 +05:30
} else {
query = fmt.Sprintf("%s GROUP BY ts_start_interval ORDER BY ts_start_interval", query)
2022-07-20 12:11:03 +05:30
}
err = r.db.Select(ctx, &logAggregatesDBResponseItems, query)
2022-07-20 12:11:03 +05:30
if err != nil {
return nil, &model.ApiError{Err: err, Typ: model.ErrorInternal}
}
aggregateResponse := model.GetLogsAggregatesResponse{
Items: make(map[int64]model.LogsAggregatesResponseItem),
}
for i := range logAggregatesDBResponseItems {
if elem, ok := aggregateResponse.Items[int64(logAggregatesDBResponseItems[i].Timestamp)]; ok {
2022-07-22 16:49:40 +05:30
if params.GroupBy != "" && logAggregatesDBResponseItems[i].GroupBy != "" {
elem.GroupBy[logAggregatesDBResponseItems[i].GroupBy] = logAggregatesDBResponseItems[i].Value
2022-07-20 12:11:03 +05:30
}
aggregateResponse.Items[logAggregatesDBResponseItems[i].Timestamp] = elem
2022-07-20 12:11:03 +05:30
} else {
2022-07-22 16:49:40 +05:30
if params.GroupBy != "" && logAggregatesDBResponseItems[i].GroupBy != "" {
aggregateResponse.Items[logAggregatesDBResponseItems[i].Timestamp] = model.LogsAggregatesResponseItem{
Timestamp: logAggregatesDBResponseItems[i].Timestamp,
GroupBy: map[string]interface{}{logAggregatesDBResponseItems[i].GroupBy: logAggregatesDBResponseItems[i].Value},
2022-07-20 12:11:03 +05:30
}
2022-07-22 16:49:40 +05:30
} else if params.GroupBy == "" {
aggregateResponse.Items[logAggregatesDBResponseItems[i].Timestamp] = model.LogsAggregatesResponseItem{
Timestamp: logAggregatesDBResponseItems[i].Timestamp,
Value: logAggregatesDBResponseItems[i].Value,
2022-07-20 12:11:03 +05:30
}
}
}
}
return &aggregateResponse, nil
}
func (r *ClickHouseReader) QueryDashboardVars(ctx context.Context, query string) (*model.DashboardVar, error) {
var result = model.DashboardVar{VariableValues: make([]interface{}, 0)}
rows, err := r.db.Query(ctx, query)
zap.L().Info(query)
if err != nil {
zap.L().Error("Error in processing sql query", zap.Error(err))
return nil, err
}
var (
columnTypes = rows.ColumnTypes()
vars = make([]interface{}, len(columnTypes))
)
for i := range columnTypes {
vars[i] = reflect.New(columnTypes[i].ScanType()).Interface()
}
defer rows.Close()
for rows.Next() {
if err := rows.Scan(vars...); err != nil {
return nil, err
}
for _, v := range vars {
switch v := v.(type) {
case *string, *int8, *int16, *int32, *int64, *uint8, *uint16, *uint32, *uint64, *float32, *float64, *time.Time, *bool:
result.VariableValues = append(result.VariableValues, reflect.ValueOf(v).Elem().Interface())
default:
return nil, fmt.Errorf("unsupported value type encountered")
}
}
}
return &result, nil
}
func (r *ClickHouseReader) GetMetricAggregateAttributes(
ctx context.Context,
req *v3.AggregateAttributeRequest,
skipDotNames bool,
) (*v3.AggregateAttributeResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.AggregateAttributeResponse
query = fmt.Sprintf("SELECT metric_name, type, is_monotonic, temporality FROM %s.%s WHERE metric_name ILIKE $1 GROUP BY metric_name, type, is_monotonic, temporality", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err = r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
seen := make(map[string]struct{})
var metricName, typ, temporality string
var isMonotonic bool
for rows.Next() {
if err := rows.Scan(&metricName, &typ, &isMonotonic, &temporality); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
if skipDotNames && strings.Contains(metricName, ".") {
continue
}
// Non-monotonic cumulative sums are treated as gauges
if typ == "Sum" && !isMonotonic && temporality == string(v3.Cumulative) {
typ = "Gauge"
}
// unlike traces/logs `tag`/`resource` type, the `Type` will be metric type
key := v3.AttributeKey{
Key: metricName,
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
DataType: v3.AttributeKeyDataTypeFloat64,
Type: v3.AttributeKeyType(typ),
IsColumn: true,
}
// remove duplicates
if _, ok := seen[metricName+typ]; ok {
continue
}
seen[metricName+typ] = struct{}{}
response.AttributeKeys = append(response.AttributeKeys, key)
}
return &response, nil
}
func (r *ClickHouseReader) GetMetricAttributeKeys(ctx context.Context, req *v3.FilterAttributeKeyRequest) (*v3.FilterAttributeKeyResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.FilterAttributeKeyResponse
// skips the internal attributes i.e attributes starting with __
query = fmt.Sprintf("SELECT arrayJoin(tagKeys) AS distinctTagKey FROM (SELECT JSONExtractKeys(labels) AS tagKeys FROM %s.%s WHERE metric_name=$1 AND unix_milli >= $2 GROUP BY tagKeys) WHERE distinctTagKey ILIKE $3 AND distinctTagKey NOT LIKE '\\_\\_%%' GROUP BY distinctTagKey", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err = r.db.Query(ctx, query, req.AggregateAttribute, common.PastDayRoundOff(), fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var attributeKey string
for rows.Next() {
if err := rows.Scan(&attributeKey); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: attributeKey,
DataType: v3.AttributeKeyDataTypeString, // https://github.com/OpenObservability/OpenMetrics/blob/main/proto/openmetrics_data_model.proto#L64-L72.
Type: v3.AttributeKeyTypeTag,
IsColumn: false,
}
response.AttributeKeys = append(response.AttributeKeys, key)
}
return &response, nil
}
func (r *ClickHouseReader) GetMetricAttributeValues(ctx context.Context, req *v3.FilterAttributeValueRequest) (*v3.FilterAttributeValueResponse, error) {
var query string
var err error
var rows driver.Rows
var attributeValues v3.FilterAttributeValueResponse
query = fmt.Sprintf("SELECT JSONExtractString(labels, $1) AS tagValue FROM %s.%s WHERE metric_name IN $2 AND JSONExtractString(labels, $3) ILIKE $4 AND unix_milli >= $5 GROUP BY tagValue", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
names := []string{req.AggregateAttribute}
if _, ok := metrics.MetricsUnderTransition[req.AggregateAttribute]; ok {
names = append(names, metrics.MetricsUnderTransition[req.AggregateAttribute])
}
rows, err = r.db.Query(ctx, query, req.FilterAttributeKey, names, req.FilterAttributeKey, fmt.Sprintf("%%%s%%", req.SearchText), common.PastDayRoundOff())
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var atrributeValue string
for rows.Next() {
if err := rows.Scan(&atrributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
// https://github.com/OpenObservability/OpenMetrics/blob/main/proto/openmetrics_data_model.proto#L64-L72
// this may change in future if we use OTLP as the data model
attributeValues.StringAttributeValues = append(attributeValues.StringAttributeValues, atrributeValue)
}
return &attributeValues, nil
}
func (r *ClickHouseReader) GetMetricMetadata(ctx context.Context, metricName, serviceName string) (*v3.MetricMetadataResponse, error) {
unixMilli := common.PastDayRoundOff()
// Note: metric metadata should be accessible regardless of the time range selection
// our standard retention period is 30 days, so we are querying the table v4_1_day to reduce the
// amount of data scanned
query := fmt.Sprintf("SELECT temporality, description, type, unit, is_monotonic from %s.%s WHERE metric_name=$1 AND unix_milli >= $2 GROUP BY temporality, description, type, unit, is_monotonic", signozMetricDBName, signozTSTableNameV41Day)
rows, err := r.db.Query(ctx, query, metricName, unixMilli)
if err != nil {
zap.L().Error("Error while fetching metric metadata", zap.Error(err))
return nil, fmt.Errorf("error while fetching metric metadata: %s", err.Error())
}
defer rows.Close()
var deltaExists, isMonotonic bool
var temporality, description, metricType, unit string
for rows.Next() {
if err := rows.Scan(&temporality, &description, &metricType, &unit, &isMonotonic); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
if temporality == string(v3.Delta) {
deltaExists = true
}
}
query = fmt.Sprintf("SELECT JSONExtractString(labels, 'le') as le from %s.%s WHERE metric_name=$1 AND unix_milli >= $2 AND type = 'Histogram' AND JSONExtractString(labels, 'service_name') = $3 GROUP BY le ORDER BY le", signozMetricDBName, signozTSTableNameV41Day)
rows, err = r.db.Query(ctx, query, metricName, unixMilli, serviceName)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var leFloat64 []float64
for rows.Next() {
var leStr string
if err := rows.Scan(&leStr); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
le, err := strconv.ParseFloat(leStr, 64)
// ignore the error and continue if the value is not a float
// ideally this should not happen but we have seen ClickHouse
// returning empty string for some values
if err != nil {
zap.L().Error("error while parsing le value", zap.Error(err))
continue
}
if math.IsInf(le, 0) {
continue
}
leFloat64 = append(leFloat64, le)
}
return &v3.MetricMetadataResponse{
Delta: deltaExists,
Le: leFloat64,
Description: description,
Unit: unit,
Type: metricType,
IsMonotonic: isMonotonic,
Temporality: temporality,
}, nil
}
// GetCountOfThings returns the count of things in the query
// This is a generic function that can be used to check if any data exists for a given query
func (r *ClickHouseReader) GetCountOfThings(ctx context.Context, query string) (uint64, error) {
var count uint64
err := r.db.QueryRow(ctx, query).Scan(&count)
if err != nil {
return 0, err
}
return count, nil
}
func (r *ClickHouseReader) GetLatestReceivedMetric(
ctx context.Context, metricNames []string, labelValues map[string]string,
) (*model.MetricStatus, *model.ApiError) {
// at least 1 metric name must be specified.
// this query can be too slow otherwise.
if len(metricNames) < 1 {
return nil, model.BadRequest(fmt.Errorf("atleast 1 metric name must be specified"))
}
quotedMetricNames := []string{}
for _, m := range metricNames {
quotedMetricNames = append(quotedMetricNames, utils.ClickHouseFormattedValue(m))
}
commaSeparatedMetricNames := strings.Join(quotedMetricNames, ", ")
whereClauseParts := []string{
fmt.Sprintf(`metric_name in (%s)`, commaSeparatedMetricNames),
}
if labelValues != nil {
for label, val := range labelValues {
whereClauseParts = append(
whereClauseParts,
fmt.Sprintf(`JSONExtractString(labels, '%s') = '%s'`, label, val),
)
}
}
if len(whereClauseParts) < 1 {
return nil, nil
}
whereClause := strings.Join(whereClauseParts, " AND ")
query := fmt.Sprintf(`
SELECT metric_name, anyLast(labels), max(unix_milli)
from %s.%s
where %s
group by metric_name
limit 1
`, signozMetricDBName, signozTSTableNameV4, whereClause,
)
rows, err := r.db.Query(ctx, query)
if err != nil {
return nil, model.InternalError(fmt.Errorf(
"couldn't query clickhouse for received metrics status: %w", err,
))
}
defer rows.Close()
var result *model.MetricStatus
if rows.Next() {
result = &model.MetricStatus{}
var labelsJson string
err := rows.Scan(
&result.MetricName,
&labelsJson,
&result.LastReceivedTsMillis,
)
if err != nil {
return nil, model.InternalError(fmt.Errorf(
"couldn't scan metric status row: %w", err,
))
}
err = json.Unmarshal([]byte(labelsJson), &result.LastReceivedLabels)
if err != nil {
return nil, model.InternalError(fmt.Errorf(
"couldn't unmarshal metric labels json: %w", err,
))
}
}
return result, nil
}
func isColumn(useLogsNewSchema bool, tableStatement, attrType, field, datType string) bool {
// value of attrType will be `resource` or `tag`, if `tag` change it to `attribute`
var name string
if useLogsNewSchema {
// adding explict '`'
name = fmt.Sprintf("`%s`", utils.GetClickhouseColumnNameV2(attrType, datType, field))
} else {
name = utils.GetClickhouseColumnName(attrType, datType, field)
}
return strings.Contains(tableStatement, fmt.Sprintf("%s ", name))
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
func (r *ClickHouseReader) GetLogAggregateAttributes(ctx context.Context, req *v3.AggregateAttributeRequest) (*v3.AggregateAttributeResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.AggregateAttributeResponse
var stringAllowed bool
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
where := ""
switch req.Operator {
case
v3.AggregateOperatorCountDistinct,
2023-04-25 21:53:46 +05:30
v3.AggregateOperatorCount:
where = "tag_key ILIKE $1"
stringAllowed = true
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
case
v3.AggregateOperatorRateSum,
v3.AggregateOperatorRateMax,
v3.AggregateOperatorRateAvg,
v3.AggregateOperatorRate,
v3.AggregateOperatorRateMin,
v3.AggregateOperatorP05,
v3.AggregateOperatorP10,
v3.AggregateOperatorP20,
v3.AggregateOperatorP25,
v3.AggregateOperatorP50,
v3.AggregateOperatorP75,
v3.AggregateOperatorP90,
v3.AggregateOperatorP95,
v3.AggregateOperatorP99,
v3.AggregateOperatorAvg,
v3.AggregateOperatorSum,
v3.AggregateOperatorMin,
v3.AggregateOperatorMax:
where = "tag_key ILIKE $1 AND (tag_data_type='int64' or tag_data_type='float64')"
stringAllowed = false
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
case
v3.AggregateOperatorNoOp:
return &v3.AggregateAttributeResponse{}, nil
default:
return nil, fmt.Errorf("unsupported aggregate operator")
}
query = fmt.Sprintf("SELECT DISTINCT(tag_key), tag_type, tag_data_type from %s.%s WHERE %s limit $2", r.logsDB, r.logsTagAttributeTableV2, where)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText), req.Limit)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.logsDB, r.logsLocalTableName)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, fmt.Errorf("error while fetching logs schema: %s", err.Error())
}
var tagKey string
var dataType string
var attType string
for rows.Next() {
if err := rows.Scan(&tagKey, &attType, &dataType); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: tagKey,
DataType: v3.AttributeKeyDataType(dataType),
Type: v3.AttributeKeyType(attType),
IsColumn: isColumn(r.useLogsNewSchema, statements[0].Statement, attType, tagKey, dataType),
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
response.AttributeKeys = append(response.AttributeKeys, key)
}
// add other attributes
for _, field := range constants.StaticFieldsLogsV3 {
if (!stringAllowed && field.DataType == v3.AttributeKeyDataTypeString) || (v3.AttributeKey{} == field) {
continue
} else if len(req.SearchText) == 0 || strings.Contains(field.Key, req.SearchText) {
response.AttributeKeys = append(response.AttributeKeys, field)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
}
return &response, nil
}
func (r *ClickHouseReader) GetLogAttributeKeys(ctx context.Context, req *v3.FilterAttributeKeyRequest) (*v3.FilterAttributeKeyResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.FilterAttributeKeyResponse
if len(req.SearchText) != 0 {
query = fmt.Sprintf("select distinct tag_key, tag_type, tag_data_type from %s.%s where tag_key ILIKE $1 limit $2", r.logsDB, r.logsTagAttributeTableV2)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText), req.Limit)
} else {
query = fmt.Sprintf("select distinct tag_key, tag_type, tag_data_type from %s.%s limit $1", r.logsDB, r.logsTagAttributeTableV2)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, req.Limit)
}
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.logsDB, r.logsLocalTableName)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, fmt.Errorf("error while fetching logs schema: %s", err.Error())
}
var attributeKey string
var attributeDataType string
var tagType string
for rows.Next() {
if err := rows.Scan(&attributeKey, &tagType, &attributeDataType); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: attributeKey,
DataType: v3.AttributeKeyDataType(attributeDataType),
Type: v3.AttributeKeyType(tagType),
IsColumn: isColumn(r.useLogsNewSchema, statements[0].Statement, tagType, attributeKey, attributeDataType),
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
response.AttributeKeys = append(response.AttributeKeys, key)
}
// add other attributes
for _, f := range constants.StaticFieldsLogsV3 {
if (v3.AttributeKey{} == f) {
continue
}
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
if len(req.SearchText) == 0 || strings.Contains(f.Key, req.SearchText) {
response.AttributeKeys = append(response.AttributeKeys, f)
}
}
return &response, nil
}
func (r *ClickHouseReader) GetLogAttributeValues(ctx context.Context, req *v3.FilterAttributeValueRequest) (*v3.FilterAttributeValueResponse, error) {
var err error
var filterValueColumn string
var rows driver.Rows
var attributeValues v3.FilterAttributeValueResponse
// if dataType or tagType is not present return empty response
if len(req.FilterAttributeKeyDataType) == 0 || len(req.TagType) == 0 {
// also check if it is not a top level key
if _, ok := constants.StaticFieldsLogsV3[req.FilterAttributeKey]; !ok {
return &v3.FilterAttributeValueResponse{}, nil
}
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
// ignore autocomplete request for body
if req.FilterAttributeKey == "body" || req.FilterAttributeKey == "__attrs" {
return &v3.FilterAttributeValueResponse{}, nil
}
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
// if data type is bool, return true and false
if req.FilterAttributeKeyDataType == v3.AttributeKeyDataTypeBool {
return &v3.FilterAttributeValueResponse{
BoolAttributeValues: []bool{true, false},
}, nil
}
query := "select distinct"
switch req.FilterAttributeKeyDataType {
case v3.AttributeKeyDataTypeInt64:
filterValueColumn = "number_value"
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
case v3.AttributeKeyDataTypeFloat64:
filterValueColumn = "number_value"
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
case v3.AttributeKeyDataTypeString:
filterValueColumn = "string_value"
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
}
searchText := fmt.Sprintf("%%%s%%", req.SearchText)
// check if the tagKey is a topLevelColumn
if _, ok := constants.StaticFieldsLogsV3[req.FilterAttributeKey]; ok {
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
// query the column for the last 48 hours
filterValueColumnWhere := req.FilterAttributeKey
selectKey := req.FilterAttributeKey
if req.FilterAttributeKeyDataType != v3.AttributeKeyDataTypeString {
filterValueColumnWhere = fmt.Sprintf("toString(%s)", req.FilterAttributeKey)
selectKey = fmt.Sprintf("toInt64(%s)", req.FilterAttributeKey)
}
// prepare the query and run
if len(req.SearchText) != 0 {
query = fmt.Sprintf("select distinct %s from %s.%s where timestamp >= toInt64(toUnixTimestamp(now() - INTERVAL 48 HOUR)*1000000000) and %s ILIKE $1 limit $2", selectKey, r.logsDB, r.logsLocalTableName, filterValueColumnWhere)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, searchText, req.Limit)
} else {
query = fmt.Sprintf("select distinct %s from %s.%s where timestamp >= toInt64(toUnixTimestamp(now() - INTERVAL 48 HOUR)*1000000000) limit $1", selectKey, r.logsDB, r.logsLocalTableName)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, req.Limit)
}
} else if len(req.SearchText) != 0 {
filterValueColumnWhere := filterValueColumn
if req.FilterAttributeKeyDataType != v3.AttributeKeyDataTypeString {
filterValueColumnWhere = fmt.Sprintf("toString(%s)", filterValueColumn)
}
query = fmt.Sprintf("SELECT DISTINCT %s FROM %s.%s WHERE tag_key=$1 AND %s ILIKE $2 AND tag_type=$3 LIMIT $4", filterValueColumn, r.logsDB, r.logsTagAttributeTableV2, filterValueColumnWhere)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, req.FilterAttributeKey, searchText, req.TagType, req.Limit)
} else {
query = fmt.Sprintf("SELECT DISTINCT %s FROM %s.%s WHERE tag_key=$1 AND tag_type=$2 LIMIT $3", filterValueColumn, r.logsDB, r.logsTagAttributeTableV2)
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
rows, err = r.db.Query(ctx, query, req.FilterAttributeKey, req.TagType, req.Limit)
}
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
feat(logs): add tag attribute autocomplete for logs (#2404) * chore: add payload types for autocomplete requests * chore: update the query params file location and payload * chore: add query range v3 API request/response payload types * feat: metric attribute autocomplete for the aggregation type * feat: add attrs filters autocomplete endpoints * feat(logs): add tag attribute autocomplete for logs * chore: added support for multiple datatype in value suggestion api for attributes * feat: int64/float64 added for AttributeKeyDataType along with validation * feat: filterAttributeValueResponse type updated * fix: number type updated and query updated * feat: remove tagType in keys autocomplete * feat: return isColumn value correctly for attibute keys * Update pkg/query-service/app/clickhouseReader/reader.go Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> * fix: don't skip empty strings in value autocomplete * fix: allow empty string search * feat: add top level column names of logs in key sugestion * fix: tagType column removed * feat: get attribute values from logs table for top level fields * feat: don't throw error if dataType and tagType is not present * feat: timerange select corrected * feat: autocomplete for int/float added * fix: reverted attributeValueResponse change * fix: null values handled for int and float * feat: add support for get log aggreagte attributes * feat: aggreate attribute logic updated and body added in keys autocomplete * fix: constants updaetd * fix: body type updated and empty response for noop and count * fix: isColumn logic updated --------- Co-authored-by: Srikanth Chekuri <srikanth.chekuri92@gmail.com> Co-authored-by: Vishal Sharma <makeavish786@gmail.com> Co-authored-by: palashgdev <palashgdev@gmail.com>
2023-04-06 13:32:24 +05:30
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var strAttributeValue string
var float64AttributeValue sql.NullFloat64
var int64AttributeValue sql.NullInt64
for rows.Next() {
switch req.FilterAttributeKeyDataType {
case v3.AttributeKeyDataTypeInt64:
if err := rows.Scan(&int64AttributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
if int64AttributeValue.Valid {
attributeValues.NumberAttributeValues = append(attributeValues.NumberAttributeValues, int64AttributeValue.Int64)
}
case v3.AttributeKeyDataTypeFloat64:
if err := rows.Scan(&float64AttributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
if float64AttributeValue.Valid {
attributeValues.NumberAttributeValues = append(attributeValues.NumberAttributeValues, float64AttributeValue.Float64)
}
case v3.AttributeKeyDataTypeString:
if err := rows.Scan(&strAttributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
attributeValues.StringAttributeValues = append(attributeValues.StringAttributeValues, strAttributeValue)
}
}
return &attributeValues, nil
}
func readRow(vars []interface{}, columnNames []string, countOfNumberCols int) ([]string, map[string]string, []map[string]string, *v3.Point) {
2023-03-23 19:45:15 +05:30
// Each row will have a value and a timestamp, and an optional list of label values
// example: {Timestamp: ..., Value: ...}
// The timestamp may also not present in some cases where the time series is reduced to single value
var point v3.Point
// groupBy is a container to hold label values for the current point
// example: ["frontend", "/fetch"]
var groupBy []string
var groupAttributesArray []map[string]string
2023-03-23 19:45:15 +05:30
// groupAttributes is a container to hold the key-value pairs for the current
// metric point.
// example: {"serviceName": "frontend", "operation": "/fetch"}
groupAttributes := make(map[string]string)
isValidPoint := false
2023-03-23 19:45:15 +05:30
for idx, v := range vars {
colName := columnNames[idx]
switch v := v.(type) {
case *string:
// special case for returning all labels in metrics datasource
if colName == "fullLabels" {
var metric map[string]string
err := json.Unmarshal([]byte(*v), &metric)
if err != nil {
zap.L().Error("unexpected error encountered", zap.Error(err))
2023-03-23 19:45:15 +05:30
}
for key, val := range metric {
groupBy = append(groupBy, val)
if _, ok := groupAttributes[key]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{key: val})
}
2023-03-23 19:45:15 +05:30
groupAttributes[key] = val
}
} else {
groupBy = append(groupBy, *v)
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: *v})
}
2023-03-23 19:45:15 +05:30
groupAttributes[colName] = *v
}
case *time.Time:
point.Timestamp = v.UnixMilli()
case *float64, *float32:
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
2023-03-23 19:45:15 +05:30
point.Value = float64(reflect.ValueOf(v).Elem().Float())
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Float()))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Float())})
}
2023-03-23 19:45:15 +05:30
groupAttributes[colName] = fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Float())
}
case **float64, **float32:
val := reflect.ValueOf(v)
if val.IsValid() && !val.IsNil() && !val.Elem().IsNil() {
value := reflect.ValueOf(v).Elem().Elem().Float()
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
point.Value = value
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", value))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", value)})
}
groupAttributes[colName] = fmt.Sprintf("%v", value)
}
}
case *uint, *uint8, *uint64, *uint16, *uint32:
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
2023-03-23 19:45:15 +05:30
point.Value = float64(reflect.ValueOf(v).Elem().Uint())
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Uint()))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Uint())})
}
2023-03-23 19:45:15 +05:30
groupAttributes[colName] = fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Uint())
}
case **uint, **uint8, **uint64, **uint16, **uint32:
val := reflect.ValueOf(v)
if val.IsValid() && !val.IsNil() && !val.Elem().IsNil() {
value := reflect.ValueOf(v).Elem().Elem().Uint()
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
point.Value = float64(value)
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", value))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", value)})
}
groupAttributes[colName] = fmt.Sprintf("%v", value)
}
}
case *int, *int8, *int16, *int32, *int64:
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
2023-03-23 19:45:15 +05:30
point.Value = float64(reflect.ValueOf(v).Elem().Int())
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Int()))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Int())})
}
2023-03-23 19:45:15 +05:30
groupAttributes[colName] = fmt.Sprintf("%v", reflect.ValueOf(v).Elem().Int())
}
case **int, **int8, **int16, **int32, **int64:
val := reflect.ValueOf(v)
if val.IsValid() && !val.IsNil() && !val.Elem().IsNil() {
value := reflect.ValueOf(v).Elem().Elem().Int()
if _, ok := constants.ReservedColumnTargetAliases[colName]; ok || countOfNumberCols == 1 {
isValidPoint = true
point.Value = float64(value)
} else {
groupBy = append(groupBy, fmt.Sprintf("%v", value))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", value)})
}
groupAttributes[colName] = fmt.Sprintf("%v", value)
}
}
case *bool:
groupBy = append(groupBy, fmt.Sprintf("%v", *v))
if _, ok := groupAttributes[colName]; !ok {
groupAttributesArray = append(groupAttributesArray, map[string]string{colName: fmt.Sprintf("%v", *v)})
}
groupAttributes[colName] = fmt.Sprintf("%v", *v)
2023-03-23 19:45:15 +05:30
default:
zap.L().Error("unsupported var type found in query builder query result", zap.Any("v", v), zap.String("colName", colName))
2023-03-23 19:45:15 +05:30
}
}
if isValidPoint {
return groupBy, groupAttributes, groupAttributesArray, &point
}
return groupBy, groupAttributes, groupAttributesArray, nil
2023-03-23 19:45:15 +05:30
}
func readRowsForTimeSeriesResult(rows driver.Rows, vars []interface{}, columnNames []string, countOfNumberCols int) ([]*v3.Series, error) {
2023-03-23 19:45:15 +05:30
// when groupBy is applied, each combination of cartesian product
// of attribute values is a separate series. Each item in seriesToPoints
// represent a unique series where the key is sorted attribute values joined
// by "," and the value is the list of points for that series
// For instance, group by (serviceName, operation)
// with two services and three operations in each will result in (maximum of) 6 series
// ("frontend", "order") x ("/fetch", "/fetch/{Id}", "/order")
//
// ("frontend", "/fetch")
// ("frontend", "/fetch/{Id}")
// ("frontend", "/order")
// ("order", "/fetch")
// ("order", "/fetch/{Id}")
// ("order", "/order")
seriesToPoints := make(map[string][]v3.Point)
var keys []string
2023-03-23 19:45:15 +05:30
// seriesToAttrs is a mapping of key to a map of attribute key to attribute value
// for each series. This is used to populate the series' attributes
// For instance, for the above example, the seriesToAttrs will be
// {
// "frontend,/fetch": {"serviceName": "frontend", "operation": "/fetch"},
// "frontend,/fetch/{Id}": {"serviceName": "frontend", "operation": "/fetch/{Id}"},
// "frontend,/order": {"serviceName": "frontend", "operation": "/order"},
// "order,/fetch": {"serviceName": "order", "operation": "/fetch"},
// "order,/fetch/{Id}": {"serviceName": "order", "operation": "/fetch/{Id}"},
// "order,/order": {"serviceName": "order", "operation": "/order"},
// }
seriesToAttrs := make(map[string]map[string]string)
labelsArray := make(map[string][]map[string]string)
2023-03-23 19:45:15 +05:30
for rows.Next() {
if err := rows.Scan(vars...); err != nil {
return nil, err
}
groupBy, groupAttributes, groupAttributesArray, metricPoint := readRow(vars, columnNames, countOfNumberCols)
2024-05-15 18:52:01 +05:30
// skip the point if the value is NaN or Inf
// are they ever useful enough to be returned?
if metricPoint != nil && (math.IsNaN(metricPoint.Value) || math.IsInf(metricPoint.Value, 0)) {
2024-05-15 18:52:01 +05:30
continue
}
2023-03-23 19:45:15 +05:30
sort.Strings(groupBy)
key := strings.Join(groupBy, "")
if _, exists := seriesToAttrs[key]; !exists {
keys = append(keys, key)
}
2023-03-23 19:45:15 +05:30
seriesToAttrs[key] = groupAttributes
labelsArray[key] = groupAttributesArray
if metricPoint != nil {
seriesToPoints[key] = append(seriesToPoints[key], *metricPoint)
}
2023-03-23 19:45:15 +05:30
}
var seriesList []*v3.Series
for _, key := range keys {
points := seriesToPoints[key]
series := v3.Series{Labels: seriesToAttrs[key], Points: points, LabelsArray: labelsArray[key]}
2023-03-23 19:45:15 +05:30
seriesList = append(seriesList, &series)
}
return seriesList, getPersonalisedError(rows.Err())
2023-03-23 19:45:15 +05:30
}
func logCommentKVs(ctx context.Context) map[string]string {
kv := ctx.Value(common.LogCommentKey)
if kv == nil {
return nil
}
logCommentKVs, ok := kv.(map[string]string)
if !ok {
return nil
}
return logCommentKVs
}
2023-03-23 19:45:15 +05:30
// GetTimeSeriesResultV3 runs the query and returns list of time series
func (r *ClickHouseReader) GetTimeSeriesResultV3(ctx context.Context, query string) ([]*v3.Series, error) {
ctxArgs := map[string]interface{}{"query": query}
for k, v := range logCommentKVs(ctx) {
ctxArgs[k] = v
}
defer utils.Elapsed("GetTimeSeriesResultV3", ctxArgs)()
2023-03-23 19:45:15 +05:30
// Hook up query progress reporting if requested.
queryId := ctx.Value("queryId")
if queryId != nil {
qid, ok := queryId.(string)
if !ok {
zap.L().Error("GetTimeSeriesResultV3: queryId in ctx not a string as expected", zap.Any("queryId", queryId))
} else {
ctx = clickhouse.Context(ctx, clickhouse.WithProgress(
func(p *clickhouse.Progress) {
go func() {
err := r.queryProgressTracker.ReportQueryProgress(qid, p)
if err != nil {
zap.L().Error(
"Couldn't report query progress",
zap.String("queryId", qid), zap.Error(err),
)
}
}()
},
))
}
}
2023-03-23 19:45:15 +05:30
rows, err := r.db.Query(ctx, query)
if err != nil {
zap.L().Error("error while reading time series result", zap.Error(err))
return nil, errors.New(err.Error())
2023-03-23 19:45:15 +05:30
}
defer rows.Close()
var (
columnTypes = rows.ColumnTypes()
columnNames = rows.Columns()
vars = make([]interface{}, len(columnTypes))
)
var countOfNumberCols int
2023-03-23 19:45:15 +05:30
for i := range columnTypes {
vars[i] = reflect.New(columnTypes[i].ScanType()).Interface()
switch columnTypes[i].ScanType().Kind() {
case reflect.Float32,
reflect.Float64,
reflect.Uint,
reflect.Uint8,
reflect.Uint16,
reflect.Uint32,
reflect.Uint64,
reflect.Int,
reflect.Int8,
reflect.Int16,
reflect.Int32,
reflect.Int64:
countOfNumberCols++
}
}
return readRowsForTimeSeriesResult(rows, vars, columnNames, countOfNumberCols)
2023-03-23 19:45:15 +05:30
}
2023-04-10 19:36:13 +05:30
// GetListResultV3 runs the query and returns list of rows
func (r *ClickHouseReader) GetListResultV3(ctx context.Context, query string) ([]*v3.Row, error) {
ctxArgs := map[string]interface{}{"query": query}
for k, v := range logCommentKVs(ctx) {
ctxArgs[k] = v
}
defer utils.Elapsed("GetListResultV3", ctxArgs)()
2023-04-10 19:36:13 +05:30
rows, err := r.db.Query(ctx, query)
if err != nil {
zap.L().Error("error while reading time series result", zap.Error(err))
return nil, errors.New(err.Error())
2023-04-10 19:36:13 +05:30
}
defer rows.Close()
var (
columnTypes = rows.ColumnTypes()
columnNames = rows.Columns()
)
var rowList []*v3.Row
for rows.Next() {
var vars = make([]interface{}, len(columnTypes))
for i := range columnTypes {
vars[i] = reflect.New(columnTypes[i].ScanType()).Interface()
}
2023-04-10 19:36:13 +05:30
if err := rows.Scan(vars...); err != nil {
return nil, err
}
row := map[string]interface{}{}
var t time.Time
for idx, v := range vars {
if columnNames[idx] == "timestamp" {
switch v := v.(type) {
case *uint64:
t = time.Unix(0, int64(*v))
case *time.Time:
t = *v
}
} else if columnNames[idx] == "timestamp_datetime" {
t = *v.(*time.Time)
} else if columnNames[idx] == "events" {
var events []map[string]interface{}
eventsFromDB, ok := v.(*[]string)
if !ok {
continue
}
for _, event := range *eventsFromDB {
var eventMap map[string]interface{}
json.Unmarshal([]byte(event), &eventMap)
events = append(events, eventMap)
}
row[columnNames[idx]] = events
} else {
row[columnNames[idx]] = v
2023-04-10 19:36:13 +05:30
}
}
// remove duplicate _ attributes for logs.
// remove this function after a month
removeDuplicateUnderscoreAttributes(row)
2023-04-10 19:36:13 +05:30
rowList = append(rowList, &v3.Row{Timestamp: t, Data: row})
}
return rowList, getPersonalisedError(rows.Err())
}
2023-04-10 19:36:13 +05:30
func getPersonalisedError(err error) error {
if err == nil {
return nil
}
zap.L().Error("error while reading result", zap.Error(err))
if strings.Contains(err.Error(), "code: 307") {
2024-05-15 18:52:01 +05:30
return chErrors.ErrResourceBytesLimitExceeded
}
if strings.Contains(err.Error(), "code: 159") {
2024-05-15 18:52:01 +05:30
return chErrors.ErrResourceTimeLimitExceeded
}
return err
2023-04-10 19:36:13 +05:30
}
func removeDuplicateUnderscoreAttributes(row map[string]interface{}) {
if val, ok := row["attributes_int64"]; ok {
attributes := val.(*map[string]int64)
for key := range *attributes {
if strings.Contains(key, ".") {
uKey := strings.ReplaceAll(key, ".", "_")
delete(*attributes, uKey)
}
}
}
if val, ok := row["attributes_float64"]; ok {
attributes := val.(*map[string]float64)
for key := range *attributes {
if strings.Contains(key, ".") {
uKey := strings.ReplaceAll(key, ".", "_")
delete(*attributes, uKey)
}
}
}
if val, ok := row["attributes_bool"]; ok {
attributes := val.(*map[string]bool)
for key := range *attributes {
if strings.Contains(key, ".") {
uKey := strings.ReplaceAll(key, ".", "_")
delete(*attributes, uKey)
}
}
}
for _, k := range []string{"attributes_string", "resources_string"} {
if val, ok := row[k]; ok {
attributes := val.(*map[string]string)
for key := range *attributes {
if strings.Contains(key, ".") {
uKey := strings.ReplaceAll(key, ".", "_")
delete(*attributes, uKey)
}
}
}
}
}
func (r *ClickHouseReader) CheckClickHouse(ctx context.Context) error {
rows, err := r.db.Query(ctx, "SELECT 1")
if err != nil {
return err
}
defer rows.Close()
return nil
}
2023-04-13 15:33:08 +05:30
func (r *ClickHouseReader) GetTraceAggregateAttributes(ctx context.Context, req *v3.AggregateAttributeRequest) (*v3.AggregateAttributeResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.AggregateAttributeResponse
var stringAllowed bool
where := ""
switch req.Operator {
case
v3.AggregateOperatorCountDistinct,
v3.AggregateOperatorCount:
where = "tag_key ILIKE $1"
stringAllowed = true
case
v3.AggregateOperatorRateSum,
v3.AggregateOperatorRateMax,
v3.AggregateOperatorRateAvg,
v3.AggregateOperatorRate,
v3.AggregateOperatorRateMin,
v3.AggregateOperatorP05,
v3.AggregateOperatorP10,
v3.AggregateOperatorP20,
v3.AggregateOperatorP25,
v3.AggregateOperatorP50,
v3.AggregateOperatorP75,
v3.AggregateOperatorP90,
v3.AggregateOperatorP95,
v3.AggregateOperatorP99,
v3.AggregateOperatorAvg,
v3.AggregateOperatorSum,
v3.AggregateOperatorMin,
v3.AggregateOperatorMax:
where = "tag_key ILIKE $1 AND tag_data_type='float64'"
stringAllowed = false
case
v3.AggregateOperatorNoOp:
return &v3.AggregateAttributeResponse{}, nil
default:
return nil, fmt.Errorf("unsupported aggregate operator")
}
query = fmt.Sprintf("SELECT DISTINCT(tag_key), tag_type, tag_data_type FROM %s.%s WHERE %s", r.TraceDB, r.spanAttributeTableV2, where)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err = r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.TraceDB, r.traceLocalTableName)
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, fmt.Errorf("error while fetching trace schema: %s", err.Error())
}
var tagKey string
var dataType string
var tagType string
for rows.Next() {
if err := rows.Scan(&tagKey, &tagType, &dataType); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: tagKey,
DataType: v3.AttributeKeyDataType(dataType),
Type: v3.AttributeKeyType(tagType),
IsColumn: isColumn(true, statements[0].Statement, tagType, tagKey, dataType),
}
if _, ok := constants.DeprecatedStaticFieldsTraces[tagKey]; !ok {
response.AttributeKeys = append(response.AttributeKeys, key)
}
}
fields := constants.NewStaticFieldsTraces
if !r.useTraceNewSchema {
fields = constants.DeprecatedStaticFieldsTraces
}
// add the new static fields
for _, field := range fields {
if (!stringAllowed && field.DataType == v3.AttributeKeyDataTypeString) || (v3.AttributeKey{} == field) {
continue
} else if len(req.SearchText) == 0 || strings.Contains(field.Key, req.SearchText) {
response.AttributeKeys = append(response.AttributeKeys, field)
}
}
return &response, nil
}
func (r *ClickHouseReader) GetTraceAttributeKeys(ctx context.Context, req *v3.FilterAttributeKeyRequest) (*v3.FilterAttributeKeyResponse, error) {
var query string
var err error
var rows driver.Rows
var response v3.FilterAttributeKeyResponse
query = fmt.Sprintf("SELECT DISTINCT(tag_key), tag_type, tag_data_type FROM %s.%s WHERE tag_key ILIKE $1 LIMIT $2", r.TraceDB, r.spanAttributeTableV2)
rows, err = r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText), req.Limit)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.TraceDB, r.traceLocalTableName)
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, fmt.Errorf("error while fetching trace schema: %s", err.Error())
}
var tagKey string
var dataType string
var tagType string
for rows.Next() {
if err := rows.Scan(&tagKey, &tagType, &dataType); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: tagKey,
DataType: v3.AttributeKeyDataType(dataType),
Type: v3.AttributeKeyType(tagType),
IsColumn: isColumn(true, statements[0].Statement, tagType, tagKey, dataType),
}
// don't send deprecated static fields
// this is added so that once the old tenants are moved to new schema,
// they old attributes are not sent to the frontend autocomplete
if _, ok := constants.DeprecatedStaticFieldsTraces[tagKey]; !ok {
response.AttributeKeys = append(response.AttributeKeys, key)
}
}
// remove this later just to have NewStaticFieldsTraces in the response
fields := constants.NewStaticFieldsTraces
if !r.useTraceNewSchema {
fields = constants.DeprecatedStaticFieldsTraces
}
// add the new static fields
for _, f := range fields {
if (v3.AttributeKey{} == f) {
continue
}
if len(req.SearchText) == 0 || strings.Contains(f.Key, req.SearchText) {
response.AttributeKeys = append(response.AttributeKeys, f)
}
}
return &response, nil
}
func (r *ClickHouseReader) GetTraceAttributeValues(ctx context.Context, req *v3.FilterAttributeValueRequest) (*v3.FilterAttributeValueResponse, error) {
var query string
var filterValueColumn string
var err error
var rows driver.Rows
var attributeValues v3.FilterAttributeValueResponse
// if dataType or tagType is not present return empty response
if len(req.FilterAttributeKeyDataType) == 0 || len(req.TagType) == 0 {
// add data type if it's a top level key
if k, ok := constants.StaticFieldsTraces[req.FilterAttributeKey]; ok {
req.FilterAttributeKeyDataType = k.DataType
} else {
return &v3.FilterAttributeValueResponse{}, nil
}
}
// if data type is bool, return true and false
if req.FilterAttributeKeyDataType == v3.AttributeKeyDataTypeBool {
return &v3.FilterAttributeValueResponse{
BoolAttributeValues: []bool{true, false},
}, nil
}
query = "SELECT DISTINCT"
switch req.FilterAttributeKeyDataType {
case v3.AttributeKeyDataTypeFloat64:
filterValueColumn = "number_value"
case v3.AttributeKeyDataTypeString:
filterValueColumn = "string_value"
2023-06-30 10:55:45 +05:30
}
searchText := fmt.Sprintf("%%%s%%", req.SearchText)
// check if the tagKey is a topLevelColumn
// here we are using StaticFieldsTraces instead of NewStaticFieldsTraces as we want to consider old columns as well.
if _, ok := constants.StaticFieldsTraces[req.FilterAttributeKey]; ok {
// query the column for the last 48 hours
filterValueColumnWhere := req.FilterAttributeKey
selectKey := req.FilterAttributeKey
if req.FilterAttributeKeyDataType != v3.AttributeKeyDataTypeString {
filterValueColumnWhere = fmt.Sprintf("toString(%s)", req.FilterAttributeKey)
selectKey = fmt.Sprintf("toInt64(%s)", req.FilterAttributeKey)
}
// TODO(nitya): remove 24 hour limit in future after checking the perf/resource implications
where := "timestamp >= toDateTime64(now() - INTERVAL 48 HOUR, 9)"
if r.useTraceNewSchema {
where += " AND ts_bucket_start >= toUInt64(toUnixTimestamp(now() - INTERVAL 48 HOUR))"
}
query = fmt.Sprintf("SELECT DISTINCT %s FROM %s.%s WHERE %s AND %s ILIKE $1 LIMIT $2", selectKey, r.TraceDB, r.traceTableName, where, filterValueColumnWhere)
rows, err = r.db.Query(ctx, query, searchText, req.Limit)
} else {
filterValueColumnWhere := filterValueColumn
if req.FilterAttributeKeyDataType != v3.AttributeKeyDataTypeString {
filterValueColumnWhere = fmt.Sprintf("toString(%s)", filterValueColumn)
}
query = fmt.Sprintf("SELECT DISTINCT %s FROM %s.%s WHERE tag_key=$1 AND %s ILIKE $2 AND tag_type=$3 LIMIT $4", filterValueColumn, r.TraceDB, r.spanAttributeTableV2, filterValueColumnWhere)
rows, err = r.db.Query(ctx, query, req.FilterAttributeKey, searchText, req.TagType, req.Limit)
}
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var strAttributeValue string
var float64AttributeValue sql.NullFloat64
for rows.Next() {
switch req.FilterAttributeKeyDataType {
case v3.AttributeKeyDataTypeFloat64:
if err := rows.Scan(&float64AttributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
if float64AttributeValue.Valid {
attributeValues.NumberAttributeValues = append(attributeValues.NumberAttributeValues, float64AttributeValue.Float64)
}
case v3.AttributeKeyDataTypeString:
if err := rows.Scan(&strAttributeValue); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
attributeValues.StringAttributeValues = append(attributeValues.StringAttributeValues, strAttributeValue)
}
}
return &attributeValues, nil
2023-06-30 10:55:45 +05:30
}
func (r *ClickHouseReader) GetSpanAttributeKeysV2(ctx context.Context) (map[string]v3.AttributeKey, error) {
var query string
var err error
var rows driver.Rows
response := map[string]v3.AttributeKey{}
query = fmt.Sprintf("SELECT DISTINCT(tagKey), tagType, dataType FROM %s.%s", r.TraceDB, r.spanAttributesKeysTable)
rows, err = r.db.Query(ctx, query)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
statements := []model.ShowCreateTableStatement{}
query = fmt.Sprintf("SHOW CREATE TABLE %s.%s", r.TraceDB, r.traceTableName)
err = r.db.Select(ctx, &statements, query)
if err != nil {
return nil, fmt.Errorf("error while fetching trace schema: %s", err.Error())
}
var tagKey string
var dataType string
var tagType string
for rows.Next() {
if err := rows.Scan(&tagKey, &tagType, &dataType); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: tagKey,
DataType: v3.AttributeKeyDataType(dataType),
Type: v3.AttributeKeyType(tagType),
IsColumn: isColumn(true, statements[0].Statement, tagType, tagKey, dataType),
}
name := tagKey + "##" + tagType + "##" + strings.ToLower(dataType)
response[name] = key
}
for _, key := range constants.StaticFieldsTraces {
name := key.Key + "##" + key.Type.String() + "##" + strings.ToLower(key.DataType.String())
response[name] = key
}
return response, nil
}
2023-04-25 21:53:46 +05:30
func (r *ClickHouseReader) GetSpanAttributeKeys(ctx context.Context) (map[string]v3.AttributeKey, error) {
if r.useTraceNewSchema {
return r.GetSpanAttributeKeysV2(ctx)
}
2023-04-25 21:53:46 +05:30
var query string
var err error
var rows driver.Rows
response := map[string]v3.AttributeKey{}
query = fmt.Sprintf("SELECT DISTINCT(tagKey), tagType, dataType, isColumn FROM %s.%s", r.TraceDB, r.spanAttributesKeysTable)
rows, err = r.db.Query(ctx, query)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
2023-04-25 21:53:46 +05:30
return nil, fmt.Errorf("error while executing query: %s", err.Error())
}
defer rows.Close()
var tagKey string
var dataType string
var tagType string
var isColumn bool
for rows.Next() {
if err := rows.Scan(&tagKey, &tagType, &dataType, &isColumn); err != nil {
return nil, fmt.Errorf("error while scanning rows: %s", err.Error())
}
key := v3.AttributeKey{
Key: tagKey,
DataType: v3.AttributeKeyDataType(dataType),
Type: v3.AttributeKeyType(tagType),
IsColumn: isColumn,
}
response[tagKey] = key
}
// add the deprecated static fields as they are not present in spanAttributeKeysTable
for _, f := range constants.DeprecatedStaticFieldsTraces {
response[f.Key] = f
}
2023-04-25 21:53:46 +05:30
return response, nil
}
func (r *ClickHouseReader) LiveTailLogsV4(ctx context.Context, query string, timestampStart uint64, idStart string, client *model.LogsLiveTailClientV2) {
if timestampStart == 0 {
timestampStart = uint64(time.Now().UnixNano())
} else {
timestampStart = uint64(utils.GetEpochNanoSecs(int64(timestampStart)))
}
ticker := time.NewTicker(time.Duration(r.liveTailRefreshSeconds) * time.Second)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
done := true
client.Done <- &done
zap.L().Debug("closing go routine : " + client.Name)
return
case <-ticker.C:
// get the new 100 logs as anything more older won't make sense
var tmpQuery string
bucketStart := (timestampStart / NANOSECOND) - 1800
// we have to form the query differently if the resource filters are used
if strings.Contains(query, r.logsResourceTableV2) {
tmpQuery = fmt.Sprintf("seen_at_ts_bucket_start >=%d)) AND ts_bucket_start >=%d AND timestamp >=%d", bucketStart, bucketStart, timestampStart)
} else {
tmpQuery = fmt.Sprintf("ts_bucket_start >=%d AND timestamp >=%d", bucketStart, timestampStart)
}
if idStart != "" {
tmpQuery = fmt.Sprintf("%s AND id > '%s'", tmpQuery, idStart)
}
// the reason we are doing desc is that we need the latest logs first
tmpQuery = query + tmpQuery + " order by timestamp desc, id desc limit 100"
// using the old structure since we can directly read it to the struct as use it.
response := []model.SignozLogV2{}
err := r.db.Select(ctx, &response, tmpQuery)
if err != nil {
zap.L().Error("Error while getting logs", zap.Error(err))
client.Error <- err
return
}
for i := len(response) - 1; i >= 0; i-- {
client.Logs <- &response[i]
if i == 0 {
timestampStart = response[i].Timestamp
idStart = response[i].ID
}
}
}
}
}
func (r *ClickHouseReader) LiveTailLogsV3(ctx context.Context, query string, timestampStart uint64, idStart string, client *model.LogsLiveTailClient) {
if timestampStart == 0 {
timestampStart = uint64(time.Now().UnixNano())
} else {
timestampStart = uint64(utils.GetEpochNanoSecs(int64(timestampStart)))
}
ticker := time.NewTicker(time.Duration(r.liveTailRefreshSeconds) * time.Second)
defer ticker.Stop()
for {
select {
case <-ctx.Done():
done := true
client.Done <- &done
zap.L().Debug("closing go routine : " + client.Name)
return
case <-ticker.C:
// get the new 100 logs as anything more older won't make sense
tmpQuery := fmt.Sprintf("timestamp >='%d'", timestampStart)
if idStart != "" {
tmpQuery = fmt.Sprintf("%s AND id > '%s'", tmpQuery, idStart)
}
// the reason we are doing desc is that we need the latest logs first
tmpQuery = query + tmpQuery + " order by timestamp desc, id desc limit 100"
// using the old structure since we can directly read it to the struct as use it.
response := []model.SignozLog{}
err := r.db.Select(ctx, &response, tmpQuery)
if err != nil {
zap.L().Error("Error while getting logs", zap.Error(err))
client.Error <- err
return
}
for i := len(response) - 1; i >= 0; i-- {
client.Logs <- &response[i]
if i == 0 {
timestampStart = response[i].Timestamp
idStart = response[i].ID
}
}
}
}
}
func (r *ClickHouseReader) AddRuleStateHistory(ctx context.Context, ruleStateHistory []model.RuleStateHistory) error {
var statement driver.Batch
var err error
defer func() {
if statement != nil {
statement.Abort()
}
}()
statement, err = r.db.PrepareBatch(ctx, fmt.Sprintf("INSERT INTO %s.%s (rule_id, rule_name, overall_state, overall_state_changed, state, state_changed, unix_milli, labels, fingerprint, value) VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10)",
signozHistoryDBName, ruleStateHistoryTableName))
if err != nil {
return err
}
for _, history := range ruleStateHistory {
err = statement.Append(history.RuleID, history.RuleName, history.OverallState, history.OverallStateChanged, history.State, history.StateChanged, history.UnixMilli, history.Labels, history.Fingerprint, history.Value)
if err != nil {
return err
}
}
err = statement.Send()
if err != nil {
return err
}
return nil
}
func (r *ClickHouseReader) GetLastSavedRuleStateHistory(ctx context.Context, ruleID string) ([]model.RuleStateHistory, error) {
query := fmt.Sprintf("SELECT * FROM %s.%s WHERE rule_id = '%s' AND state_changed = true ORDER BY unix_milli DESC LIMIT 1 BY fingerprint",
signozHistoryDBName, ruleStateHistoryTableName, ruleID)
history := []model.RuleStateHistory{}
err := r.db.Select(ctx, &history, query)
if err != nil {
return nil, err
}
return history, nil
}
func (r *ClickHouseReader) ReadRuleStateHistoryByRuleID(
ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) (*model.RuleStateTimeline, error) {
var conditions []string
conditions = append(conditions, fmt.Sprintf("rule_id = '%s'", ruleID))
conditions = append(conditions, fmt.Sprintf("unix_milli >= %d AND unix_milli < %d", params.Start, params.End))
if params.State != "" {
conditions = append(conditions, fmt.Sprintf("state = '%s'", params.State))
}
if params.Filters != nil && len(params.Filters.Items) != 0 {
for _, item := range params.Filters.Items {
toFormat := item.Value
op := v3.FilterOperator(strings.ToLower(strings.TrimSpace(string(item.Operator))))
if op == v3.FilterOperatorContains || op == v3.FilterOperatorNotContains {
toFormat = fmt.Sprintf("%%%s%%", toFormat)
}
fmtVal := utils.ClickHouseFormattedValue(toFormat)
switch op {
case v3.FilterOperatorEqual:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') = %s", item.Key.Key, fmtVal))
case v3.FilterOperatorNotEqual:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') != %s", item.Key.Key, fmtVal))
case v3.FilterOperatorIn:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') IN %s", item.Key.Key, fmtVal))
case v3.FilterOperatorNotIn:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') NOT IN %s", item.Key.Key, fmtVal))
case v3.FilterOperatorLike:
conditions = append(conditions, fmt.Sprintf("like(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorNotLike:
conditions = append(conditions, fmt.Sprintf("notLike(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorRegex:
conditions = append(conditions, fmt.Sprintf("match(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorNotRegex:
conditions = append(conditions, fmt.Sprintf("not match(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorGreaterThan:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') > %s", item.Key.Key, fmtVal))
case v3.FilterOperatorGreaterThanOrEq:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') >= %s", item.Key.Key, fmtVal))
case v3.FilterOperatorLessThan:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') < %s", item.Key.Key, fmtVal))
case v3.FilterOperatorLessThanOrEq:
conditions = append(conditions, fmt.Sprintf("JSONExtractString(labels, '%s') <= %s", item.Key.Key, fmtVal))
case v3.FilterOperatorContains:
conditions = append(conditions, fmt.Sprintf("like(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorNotContains:
conditions = append(conditions, fmt.Sprintf("notLike(JSONExtractString(labels, '%s'), %s)", item.Key.Key, fmtVal))
case v3.FilterOperatorExists:
conditions = append(conditions, fmt.Sprintf("has(JSONExtractKeys(labels), '%s')", item.Key.Key))
case v3.FilterOperatorNotExists:
conditions = append(conditions, fmt.Sprintf("not has(JSONExtractKeys(labels), '%s')", item.Key.Key))
default:
return nil, fmt.Errorf("unsupported filter operator")
}
}
}
whereClause := strings.Join(conditions, " AND ")
query := fmt.Sprintf("SELECT * FROM %s.%s WHERE %s ORDER BY unix_milli %s LIMIT %d OFFSET %d",
signozHistoryDBName, ruleStateHistoryTableName, whereClause, params.Order, params.Limit, params.Offset)
history := []model.RuleStateHistory{}
zap.L().Debug("rule state history query", zap.String("query", query))
err := r.db.Select(ctx, &history, query)
if err != nil {
zap.L().Error("Error while reading rule state history", zap.Error(err))
return nil, err
}
var total uint64
zap.L().Debug("rule state history total query", zap.String("query", fmt.Sprintf("SELECT count(*) FROM %s.%s WHERE %s",
signozHistoryDBName, ruleStateHistoryTableName, whereClause)))
err = r.db.QueryRow(ctx, fmt.Sprintf("SELECT count(*) FROM %s.%s WHERE %s",
signozHistoryDBName, ruleStateHistoryTableName, whereClause)).Scan(&total)
if err != nil {
return nil, err
}
labelsQuery := fmt.Sprintf("SELECT DISTINCT labels FROM %s.%s WHERE rule_id = $1",
signozHistoryDBName, ruleStateHistoryTableName)
rows, err := r.db.Query(ctx, labelsQuery, ruleID)
if err != nil {
return nil, err
}
defer rows.Close()
labelsMap := make(map[string][]string)
for rows.Next() {
var rawLabel string
err = rows.Scan(&rawLabel)
if err != nil {
return nil, err
}
label := map[string]string{}
err = json.Unmarshal([]byte(rawLabel), &label)
if err != nil {
return nil, err
}
for k, v := range label {
labelsMap[k] = append(labelsMap[k], v)
}
}
timeline := &model.RuleStateTimeline{
Items: history,
Total: total,
Labels: labelsMap,
}
return timeline, nil
}
func (r *ClickHouseReader) ReadRuleStateHistoryTopContributorsByRuleID(
ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) ([]model.RuleStateHistoryContributor, error) {
query := fmt.Sprintf(`SELECT
fingerprint,
any(labels) as labels,
count(*) as count
FROM %s.%s
WHERE rule_id = '%s' AND (state_changed = true) AND (state = '%s') AND unix_milli >= %d AND unix_milli <= %d
GROUP BY fingerprint
HAVING labels != '{}'
ORDER BY count DESC`,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, model.StateFiring.String(), params.Start, params.End)
zap.L().Debug("rule state history top contributors query", zap.String("query", query))
contributors := []model.RuleStateHistoryContributor{}
err := r.db.Select(ctx, &contributors, query)
if err != nil {
zap.L().Error("Error while reading rule state history", zap.Error(err))
return nil, err
}
return contributors, nil
}
func (r *ClickHouseReader) GetOverallStateTransitions(ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) ([]model.ReleStateItem, error) {
tmpl := `WITH firing_events AS (
SELECT
rule_id,
state,
unix_milli AS firing_time
FROM %s.%s
WHERE overall_state = '` + model.StateFiring.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
resolution_events AS (
SELECT
rule_id,
state,
unix_milli AS resolution_time
FROM %s.%s
WHERE overall_state = '` + model.StateInactive.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
matched_events AS (
SELECT
f.rule_id,
f.state,
f.firing_time,
MIN(r.resolution_time) AS resolution_time
FROM firing_events f
LEFT JOIN resolution_events r
ON f.rule_id = r.rule_id
WHERE r.resolution_time > f.firing_time
GROUP BY f.rule_id, f.state, f.firing_time
)
SELECT *
FROM matched_events
ORDER BY firing_time ASC;`
query := fmt.Sprintf(tmpl,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End)
zap.L().Debug("overall state transitions query", zap.String("query", query))
transitions := []model.RuleStateTransition{}
err := r.db.Select(ctx, &transitions, query)
if err != nil {
return nil, err
}
stateItems := []model.ReleStateItem{}
for idx, item := range transitions {
start := item.FiringTime
end := item.ResolutionTime
stateItems = append(stateItems, model.ReleStateItem{
State: item.State,
Start: start,
End: end,
})
if idx < len(transitions)-1 {
nextStart := transitions[idx+1].FiringTime
if nextStart > end {
stateItems = append(stateItems, model.ReleStateItem{
State: model.StateInactive,
Start: end,
End: nextStart,
})
}
}
}
// fetch the most recent overall_state from the table
var state model.AlertState
stateQuery := fmt.Sprintf("SELECT state FROM %s.%s WHERE rule_id = '%s' AND unix_milli <= %d ORDER BY unix_milli DESC LIMIT 1",
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.End)
if err := r.db.QueryRow(ctx, stateQuery).Scan(&state); err != nil {
if err != sql.ErrNoRows {
return nil, err
}
state = model.StateInactive
}
if len(transitions) == 0 {
// no transitions found, it is either firing or inactive for whole time range
stateItems = append(stateItems, model.ReleStateItem{
State: state,
Start: params.Start,
End: params.End,
})
} else {
// there were some transitions, we need to add the last state at the end
if state == model.StateInactive {
stateItems = append(stateItems, model.ReleStateItem{
State: model.StateInactive,
Start: transitions[len(transitions)-1].ResolutionTime,
End: params.End,
})
} else {
// fetch the most recent firing event from the table in the given time range
var firingTime int64
firingQuery := fmt.Sprintf(`
SELECT
unix_milli
FROM %s.%s
WHERE rule_id = '%s' AND overall_state_changed = true AND overall_state = '%s' AND unix_milli <= %d
ORDER BY unix_milli DESC LIMIT 1`, signozHistoryDBName, ruleStateHistoryTableName, ruleID, model.StateFiring.String(), params.End)
if err := r.db.QueryRow(ctx, firingQuery).Scan(&firingTime); err != nil {
return nil, err
}
stateItems = append(stateItems, model.ReleStateItem{
State: model.StateInactive,
Start: transitions[len(transitions)-1].ResolutionTime,
End: firingTime,
})
stateItems = append(stateItems, model.ReleStateItem{
State: model.StateFiring,
Start: firingTime,
End: params.End,
})
}
}
return stateItems, nil
}
func (r *ClickHouseReader) GetAvgResolutionTime(ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) (float64, error) {
tmpl := `
WITH firing_events AS (
SELECT
rule_id,
state,
unix_milli AS firing_time
FROM %s.%s
WHERE overall_state = '` + model.StateFiring.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
resolution_events AS (
SELECT
rule_id,
state,
unix_milli AS resolution_time
FROM %s.%s
WHERE overall_state = '` + model.StateInactive.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
matched_events AS (
SELECT
f.rule_id,
f.state,
f.firing_time,
MIN(r.resolution_time) AS resolution_time
FROM firing_events f
LEFT JOIN resolution_events r
ON f.rule_id = r.rule_id
WHERE r.resolution_time > f.firing_time
GROUP BY f.rule_id, f.state, f.firing_time
)
SELECT AVG(resolution_time - firing_time) / 1000 AS avg_resolution_time
FROM matched_events;
`
query := fmt.Sprintf(tmpl,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End)
zap.L().Debug("avg resolution time query", zap.String("query", query))
var avgResolutionTime float64
err := r.db.QueryRow(ctx, query).Scan(&avgResolutionTime)
if err != nil {
return 0, err
}
return avgResolutionTime, nil
}
func (r *ClickHouseReader) GetAvgResolutionTimeByInterval(ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) (*v3.Series, error) {
step := common.MinAllowedStepInterval(params.Start, params.End)
tmpl := `
WITH firing_events AS (
SELECT
rule_id,
state,
unix_milli AS firing_time
FROM %s.%s
WHERE overall_state = '` + model.StateFiring.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
resolution_events AS (
SELECT
rule_id,
state,
unix_milli AS resolution_time
FROM %s.%s
WHERE overall_state = '` + model.StateInactive.String() + `'
AND overall_state_changed = true
AND rule_id IN ('%s')
AND unix_milli >= %d AND unix_milli <= %d
),
matched_events AS (
SELECT
f.rule_id,
f.state,
f.firing_time,
MIN(r.resolution_time) AS resolution_time
FROM firing_events f
LEFT JOIN resolution_events r
ON f.rule_id = r.rule_id
WHERE r.resolution_time > f.firing_time
GROUP BY f.rule_id, f.state, f.firing_time
)
SELECT toStartOfInterval(toDateTime(firing_time / 1000), INTERVAL %d SECOND) AS ts, AVG(resolution_time - firing_time) / 1000 AS avg_resolution_time
FROM matched_events
GROUP BY ts
ORDER BY ts ASC;`
query := fmt.Sprintf(tmpl,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End,
signozHistoryDBName, ruleStateHistoryTableName, ruleID, params.Start, params.End, step)
zap.L().Debug("avg resolution time by interval query", zap.String("query", query))
result, err := r.GetTimeSeriesResultV3(ctx, query)
if err != nil || len(result) == 0 {
return nil, err
}
return result[0], nil
}
func (r *ClickHouseReader) GetTotalTriggers(ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) (uint64, error) {
query := fmt.Sprintf("SELECT count(*) FROM %s.%s WHERE rule_id = '%s' AND (state_changed = true) AND (state = '%s') AND unix_milli >= %d AND unix_milli <= %d",
signozHistoryDBName, ruleStateHistoryTableName, ruleID, model.StateFiring.String(), params.Start, params.End)
var totalTriggers uint64
err := r.db.QueryRow(ctx, query).Scan(&totalTriggers)
if err != nil {
return 0, err
}
return totalTriggers, nil
}
func (r *ClickHouseReader) GetTriggersByInterval(ctx context.Context, ruleID string, params *model.QueryRuleStateHistory) (*v3.Series, error) {
step := common.MinAllowedStepInterval(params.Start, params.End)
query := fmt.Sprintf("SELECT count(*), toStartOfInterval(toDateTime(intDiv(unix_milli, 1000)), INTERVAL %d SECOND) as ts FROM %s.%s WHERE rule_id = '%s' AND (state_changed = true) AND (state = '%s') AND unix_milli >= %d AND unix_milli <= %d GROUP BY ts ORDER BY ts ASC",
step, signozHistoryDBName, ruleStateHistoryTableName, ruleID, model.StateFiring.String(), params.Start, params.End)
result, err := r.GetTimeSeriesResultV3(ctx, query)
if err != nil || len(result) == 0 {
return nil, err
}
return result[0], nil
}
func (r *ClickHouseReader) GetMinAndMaxTimestampForTraceID(ctx context.Context, traceID []string) (int64, int64, error) {
var minTime, maxTime time.Time
query := fmt.Sprintf("SELECT min(timestamp), max(timestamp) FROM %s.%s WHERE traceID IN ('%s')",
r.TraceDB, r.SpansTable, strings.Join(traceID, "','"))
zap.L().Debug("GetMinAndMaxTimestampForTraceID", zap.String("query", query))
err := r.db.QueryRow(ctx, query).Scan(&minTime, &maxTime)
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return 0, 0, err
}
// return current time if traceID not found
if minTime.IsZero() || maxTime.IsZero() {
zap.L().Debug("minTime or maxTime is zero, traceID not found")
return time.Now().UnixNano(), time.Now().UnixNano(), nil
}
zap.L().Debug("GetMinAndMaxTimestampForTraceID", zap.Any("minTime", minTime), zap.Any("maxTime", maxTime))
return minTime.UnixNano(), maxTime.UnixNano(), nil
}
func (r *ClickHouseReader) ReportQueryStartForProgressTracking(
queryId string,
) (func(), *model.ApiError) {
return r.queryProgressTracker.ReportQueryStarted(queryId)
}
func (r *ClickHouseReader) SubscribeToQueryProgress(
queryId string,
) (<-chan model.QueryProgress, func(), *model.ApiError) {
return r.queryProgressTracker.SubscribeToQueryProgress(queryId)
}
func (r *ClickHouseReader) GetAllMetricFilterAttributeKeys(ctx context.Context, req *metrics_explorer.FilterKeyRequest, skipDotNames bool) (*[]v3.AttributeKey, *model.ApiError) {
var rows driver.Rows
var response []v3.AttributeKey
query := fmt.Sprintf("SELECT arrayJoin(tagKeys) AS distinctTagKey FROM (SELECT JSONExtractKeys(labels) AS tagKeys FROM %s.%s WHERE unix_milli >= $1 GROUP BY tagKeys) WHERE distinctTagKey ILIKE $2 AND distinctTagKey NOT LIKE '\\_\\_%%' GROUP BY distinctTagKey", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err := r.db.Query(ctx, query, common.PastDayRoundOff(), fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
var attributeKey string
for rows.Next() {
if err := rows.Scan(&attributeKey); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
if skipDotNames && strings.Contains(attributeKey, ".") {
continue
}
key := v3.AttributeKey{
Key: attributeKey,
DataType: v3.AttributeKeyDataTypeString, // https://github.com/OpenObservability/OpenMetrics/blob/main/proto/openmetrics_data_model.proto#L64-L72.
Type: v3.AttributeKeyTypeTag,
IsColumn: false,
}
response = append(response, key)
}
return &response, nil
}
func (r *ClickHouseReader) GetAllMetricFilterAttributeValues(ctx context.Context, req *metrics_explorer.FilterValueRequest) ([]string, *model.ApiError) {
var query string
var err error
var rows driver.Rows
var attributeValues []string
query = fmt.Sprintf("SELECT JSONExtractString(labels, $1) AS tagValue FROM %s.%s WHERE JSONExtractString(labels, $2) ILIKE $3 AND unix_milli >= $4 GROUP BY tagValue", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err = r.db.Query(ctx, query, req.FilterKey, req.FilterKey, fmt.Sprintf("%%%s%%", req.SearchText), common.PastDayRoundOff())
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
var atrributeValue string
for rows.Next() {
if err := rows.Scan(&atrributeValue); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
attributeValues = append(attributeValues, atrributeValue)
}
return attributeValues, nil
}
func (r *ClickHouseReader) GetAllMetricFilterUnits(ctx context.Context, req *metrics_explorer.FilterValueRequest) ([]string, *model.ApiError) {
var rows driver.Rows
var response []string
query := fmt.Sprintf("SELECT DISTINCT unit FROM %s.%s WHERE unit ILIKE $1 AND unit IS NOT NULL ORDER BY unit", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err := r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
var attributeKey string
for rows.Next() {
if err := rows.Scan(&attributeKey); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
response = append(response, attributeKey)
}
return response, nil
}
func (r *ClickHouseReader) GetAllMetricFilterTypes(ctx context.Context, req *metrics_explorer.FilterValueRequest) ([]string, *model.ApiError) {
var rows driver.Rows
var response []string
query := fmt.Sprintf("SELECT DISTINCT type FROM %s.%s WHERE type ILIKE $1 AND type IS NOT NULL ORDER BY type", signozMetricDBName, signozTSTableNameV41Day)
if req.Limit != 0 {
query = query + fmt.Sprintf(" LIMIT %d;", req.Limit)
}
rows, err := r.db.Query(ctx, query, fmt.Sprintf("%%%s%%", req.SearchText))
if err != nil {
zap.L().Error("Error while executing query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
var attributeKey string
for rows.Next() {
if err := rows.Scan(&attributeKey); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
response = append(response, attributeKey)
}
return response, nil
}
func (r *ClickHouseReader) GetMetricsDataPoints(ctx context.Context, metricName string) (uint64, *model.ApiError) {
query := fmt.Sprintf(`SELECT
sum(count) as data_points
FROM %s.%s
WHERE metric_name = ?
`, signozMetricDBName, constants.SIGNOZ_SAMPLES_V4_AGG_30M_TABLENAME)
var dataPoints uint64
err := r.db.QueryRow(ctx, query, metricName).Scan(&dataPoints)
if err != nil {
return 0, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return dataPoints, nil // Convert to uint64 before returning
}
func (r *ClickHouseReader) GetMetricsLastReceived(ctx context.Context, metricName string) (int64, *model.ApiError) {
query := fmt.Sprintf(`SELECT
MAX(unix_milli) AS last_received_time
FROM %s.%s
WHERE metric_name = ?
`, signozMetricDBName, signozSampleTableName)
var lastReceived int64
err := r.db.QueryRow(ctx, query, metricName).Scan(&lastReceived)
if err != nil {
return 0, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return lastReceived, nil // Convert to uint64 before returning
}
func (r *ClickHouseReader) GetTotalTimeSeriesForMetricName(ctx context.Context, metricName string) (uint64, *model.ApiError) {
query := fmt.Sprintf(`SELECT
uniq(fingerprint) AS timeSeriesCount
FROM %s.%s
WHERE metric_name = ?;`, signozMetricDBName, signozTSTableNameV41Week)
var timeSeriesCount uint64
err := r.db.QueryRow(ctx, query, metricName).Scan(&timeSeriesCount)
if err != nil {
return 0, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return timeSeriesCount, nil
}
func (r *ClickHouseReader) GetAttributesForMetricName(ctx context.Context, metricName string) (*[]metrics_explorer.Attribute, *model.ApiError) {
query := fmt.Sprintf(`
SELECT
kv.1 AS key,
arrayMap(x -> trim(BOTH '\"' FROM x), groupUniqArray(10000)(kv.2)) AS values,
length(groupUniqArray(10000)(kv.2)) AS valueCount
FROM %s.%s
ARRAY JOIN arrayFilter(x -> NOT startsWith(x.1, '__'), JSONExtractKeysAndValuesRaw(labels)) AS kv
WHERE metric_name = ?
GROUP BY kv.1
ORDER BY valueCount DESC;
`, signozMetricDBName, signozTSTableNameV41Week)
rows, err := r.db.Query(ctx, query, metricName)
if err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close() // Ensure the rows are closed
var attributesList []metrics_explorer.Attribute
for rows.Next() {
var key string
var values []string
var valueCount uint64
// Manually scan each value into its corresponding variable
if err := rows.Scan(&key, &values, &valueCount); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
// Append the scanned values into the struct
attributesList = append(attributesList, metrics_explorer.Attribute{
Key: key,
Value: values,
ValueCount: valueCount,
})
}
// Handle any errors encountered while scanning rows
if err := rows.Err(); err != nil {
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return &attributesList, nil
}
func (r *ClickHouseReader) GetActiveTimeSeriesForMetricName(ctx context.Context, metricName string, duration time.Duration) (uint64, *model.ApiError) {
milli := time.Now().Add(-duration).UnixMilli()
query := fmt.Sprintf("SELECT uniq(fingerprint) FROM %s.%s WHERE metric_name = '%s' and unix_milli >= ?", signozMetricDBName, signozTSTableNameV4, metricName)
var timeSeries uint64
// Using QueryRow instead of Select since we're only expecting a single value
err := r.db.QueryRow(ctx, query, milli).Scan(&timeSeries)
if err != nil {
return 0, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return timeSeries, nil
}
func (r *ClickHouseReader) ListSummaryMetrics(ctx context.Context, req *metrics_explorer.SummaryListMetricsRequest) (*metrics_explorer.SummaryListMetricsResponse, *model.ApiError) {
var args []interface{}
conditions, _ := utils.BuildFilterConditions(&req.Filters, "t")
whereClause := ""
if conditions != nil {
whereClause = "AND " + strings.Join(conditions, " AND ")
}
firstQueryLimit := req.Limit
dataPointsOrder := false
var orderByClauseFirstQuery string
if req.OrderBy.ColumnName == "samples" {
dataPointsOrder = true
orderByClauseFirstQuery = fmt.Sprintf("ORDER BY timeseries %s", req.OrderBy.Order)
if req.Limit < 50 {
firstQueryLimit = 50
}
} else if req.OrderBy.ColumnName == "metric_type" {
orderByClauseFirstQuery = fmt.Sprintf("ORDER BY type %s", req.OrderBy.Order)
} else {
orderByClauseFirstQuery = fmt.Sprintf("ORDER BY %s %s", req.OrderBy.ColumnName, req.OrderBy.Order)
}
start, end, tsTable, localTsTable := utils.WhichTSTableToUse(req.Start, req.EndD)
sampleTable, countExp := utils.WhichSampleTableToUse(req.Start, req.EndD)
metricsQuery := fmt.Sprintf(
`SELECT
t.metric_name AS metric_name,
ANY_VALUE(t.description) AS description,
ANY_VALUE(t.type) AS type,
ANY_VALUE(t.unit),
uniq(t.fingerprint) AS timeseries,
uniq(metric_name) OVER() AS total
FROM %s.%s AS t
WHERE unix_milli BETWEEN ? AND ?
%s
GROUP BY t.metric_name
%s
LIMIT %d OFFSET %d;`,
signozMetricDBName, tsTable, whereClause, orderByClauseFirstQuery, firstQueryLimit, req.Offset)
args = append(args, start, end)
valueCtx := context.WithValue(ctx, "clickhouse_max_threads", constants.MetricsExplorerClickhouseThreads)
rows, err := r.db.Query(valueCtx, metricsQuery, args...)
if err != nil {
zap.L().Error("Error executing metrics query", zap.Error(err))
return &metrics_explorer.SummaryListMetricsResponse{}, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
var response metrics_explorer.SummaryListMetricsResponse
var metricNames []string
for rows.Next() {
var metric metrics_explorer.MetricDetail
if err := rows.Scan(&metric.MetricName, &metric.Description, &metric.Type, &metric.Unit, &metric.TimeSeries, &response.Total); err != nil {
zap.L().Error("Error scanning metric row", zap.Error(err))
return &response, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
metricNames = append(metricNames, metric.MetricName)
response.Metrics = append(response.Metrics, metric)
}
if err := rows.Err(); err != nil {
zap.L().Error("Error iterating over metric rows", zap.Error(err))
return &response, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
if len(metricNames) == 0 {
return &response, nil
}
metricsList := "'" + strings.Join(metricNames, "', '") + "'"
if dataPointsOrder {
orderByClauseFirstQuery = fmt.Sprintf("ORDER BY s.samples %s", req.OrderBy.Order)
} else {
orderByClauseFirstQuery = ""
}
var sb strings.Builder
sb.WriteString(fmt.Sprintf(
`SELECT
s.samples,
s.metric_name,
s.unix_milli AS lastReceived
FROM (
SELECT
metric_name,
%s AS samples,
max(unix_milli) as unix_milli
FROM %s.%s
`, countExp, signozMetricDBName, sampleTable))
// Conditionally add the fingerprint subquery if `whereClause` is present
if whereClause != "" {
sb.WriteString(fmt.Sprintf(
`WHERE fingerprint IN (
SELECT fingerprint
FROM %s.%s
WHERE unix_milli BETWEEN ? AND ?
%s
AND metric_name IN (%s)
GROUP BY fingerprint
)
AND metric_name IN (%s) `,
signozMetricDBName, localTsTable, whereClause, metricsList, metricsList))
} else {
sb.WriteString(fmt.Sprintf(
`WHERE metric_name IN (%s) `, metricsList))
}
sb.WriteString(`GROUP BY metric_name ) AS s `)
if orderByClauseFirstQuery != "" {
sb.WriteString(orderByClauseFirstQuery)
}
sb.WriteString(fmt.Sprintf(" LIMIT %d OFFSET %d;", req.Limit, req.Offset))
sampleQuery := sb.String()
args = append(args, start, end)
rows, err = r.db.Query(valueCtx, sampleQuery, args...)
if err != nil {
zap.L().Error("Error executing samples query", zap.Error(err))
return &response, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
samplesMap := make(map[string]uint64)
lastReceivedMap := make(map[string]int64)
for rows.Next() {
var samples uint64
var metricName string
var lastReceived int64
if err := rows.Scan(&samples, &metricName, &lastReceived); err != nil {
zap.L().Error("Error scanning sample row", zap.Error(err))
return &response, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
samplesMap[metricName] = samples
lastReceivedMap[metricName] = lastReceived
}
if err := rows.Err(); err != nil {
zap.L().Error("Error iterating over sample rows", zap.Error(err))
return &response, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
var filteredMetrics []metrics_explorer.MetricDetail
for i := range response.Metrics {
if samples, exists := samplesMap[response.Metrics[i].MetricName]; exists {
response.Metrics[i].Samples = samples
if lastReceived, exists := lastReceivedMap[response.Metrics[i].MetricName]; exists {
response.Metrics[i].LastReceived = lastReceived
}
filteredMetrics = append(filteredMetrics, response.Metrics[i])
}
}
response.Metrics = filteredMetrics
if dataPointsOrder {
sort.Slice(response.Metrics, func(i, j int) bool {
return response.Metrics[i].Samples > response.Metrics[j].Samples
})
}
return &response, nil
}
func (r *ClickHouseReader) GetMetricsTimeSeriesPercentage(ctx context.Context, req *metrics_explorer.TreeMapMetricsRequest) (*[]metrics_explorer.TreeMapResponseItem, *model.ApiError) {
var args []interface{}
// Build filters dynamically
conditions, _ := utils.BuildFilterConditions(&req.Filters, "")
whereClause := ""
if len(conditions) > 0 {
whereClause = "AND " + strings.Join(conditions, " AND ")
}
start, end, tsTable, _ := utils.WhichTSTableToUse(req.Start, req.EndD)
// Construct the query without backticks
query := fmt.Sprintf(`
SELECT
metric_name,
total_value,
(total_value * 100.0 / total_time_series) AS percentage
FROM (
SELECT
metric_name,
uniq(fingerprint) AS total_value,
(SELECT uniq(fingerprint)
FROM %s.%s
WHERE unix_milli BETWEEN ? AND ?) AS total_time_series
FROM %s.%s
WHERE unix_milli BETWEEN ? AND ? %s
GROUP BY metric_name
)
ORDER BY percentage DESC
LIMIT %d;`,
signozMetricDBName,
tsTable,
signozMetricDBName,
tsTable,
whereClause,
req.Limit,
)
args = append(args,
start, end, // For total_cardinality subquery
start, end, // For main query
)
valueCtx := context.WithValue(ctx, "clickhouse_max_threads", constants.MetricsExplorerClickhouseThreads)
rows, err := r.db.Query(valueCtx, query, args...)
if err != nil {
zap.L().Error("Error executing cardinality query", zap.Error(err), zap.String("query", query))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
var heatmap []metrics_explorer.TreeMapResponseItem
for rows.Next() {
var item metrics_explorer.TreeMapResponseItem
if err := rows.Scan(&item.MetricName, &item.TotalValue, &item.Percentage); err != nil {
zap.L().Error("Error scanning row", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
heatmap = append(heatmap, item)
}
if err := rows.Err(); err != nil {
zap.L().Error("Error iterating over rows", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return &heatmap, nil
}
func (r *ClickHouseReader) GetMetricsSamplesPercentage(ctx context.Context, req *metrics_explorer.TreeMapMetricsRequest) (*[]metrics_explorer.TreeMapResponseItem, *model.ApiError) {
var args []interface{}
// Build the filter conditions
conditions, _ := utils.BuildFilterConditions(&req.Filters, "t")
whereClause := ""
if conditions != nil {
whereClause = "AND " + strings.Join(conditions, " AND ")
}
// Determine time range and tables to use
start, end, tsTable, localTsTable := utils.WhichTSTableToUse(req.Start, req.EndD)
sampleTable, countExp := utils.WhichSampleTableToUse(req.Start, req.EndD)
// Construct the metrics query
queryLimit := 50 + req.Limit
metricsQuery := fmt.Sprintf(
`SELECT
t.metric_name AS metric_name,
uniq(t.fingerprint) AS timeSeries
FROM %s.%s AS t
WHERE unix_milli BETWEEN ? AND ?
%s
GROUP BY t.metric_name
ORDER BY timeSeries DESC
LIMIT %d;`,
signozMetricDBName, tsTable, whereClause, queryLimit,
)
args = append(args, start, end)
valueCtx := context.WithValue(ctx, "clickhouse_max_threads", constants.MetricsExplorerClickhouseThreads)
// Execute the metrics query
rows, err := r.db.Query(valueCtx, metricsQuery, args...)
if err != nil {
zap.L().Error("Error executing metrics query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
// Process the query results
var metricNames []string
for rows.Next() {
var metricName string
var timeSeries uint64
if err := rows.Scan(&metricName, &timeSeries); err != nil {
zap.L().Error("Error scanning metric row", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
metricNames = append(metricNames, metricName)
}
if err := rows.Err(); err != nil {
zap.L().Error("Error iterating over metric rows", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
// If no metrics found, return early
if len(metricNames) == 0 {
return nil, nil
}
// Format metric names for query
metricsList := "'" + strings.Join(metricNames, "', '") + "'"
// Build query using string builder for better performance
var sb strings.Builder
sb.WriteString(fmt.Sprintf(
`WITH TotalSamples AS (
SELECT %s AS total_samples
FROM %s.%s
WHERE unix_milli BETWEEN ? AND ?
)
SELECT
s.samples,
s.metric_name,
COALESCE((s.samples * 100.0 / t.total_samples), 0) AS percentage
FROM
(
SELECT
metric_name,
%s AS samples
FROM %s.%s`,
countExp, signozMetricDBName, sampleTable, // Total samples
countExp, signozMetricDBName, sampleTable, // Inner select samples
))
// Conditionally add the fingerprint subquery if whereClause is present
if whereClause != "" {
sb.WriteString(fmt.Sprintf(
` WHERE fingerprint IN (
SELECT fingerprint
FROM %s.%s
WHERE unix_milli BETWEEN ? AND ?
%s
AND metric_name IN (%s)
GROUP BY fingerprint
)
AND metric_name IN (%s)`,
signozMetricDBName, localTsTable, whereClause, metricsList,
metricsList,
))
} else {
sb.WriteString(fmt.Sprintf(
` WHERE metric_name IN (%s)`,
metricsList,
))
}
sb.WriteString(`
GROUP BY metric_name
) AS s
JOIN TotalSamples t ON 1 = 1
ORDER BY percentage DESC
LIMIT ?;`)
sampleQuery := sb.String()
// Add start and end time to args
args = append(args, start, end)
// Execute the sample percentage query
rows, err = r.db.Query(valueCtx, sampleQuery, args...)
if err != nil {
zap.L().Error("Error executing samples query", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
defer rows.Close()
// Process the results into a response slice
var heatmap []metrics_explorer.TreeMapResponseItem
for rows.Next() {
var item metrics_explorer.TreeMapResponseItem
if err := rows.Scan(&item.TotalValue, &item.MetricName, &item.Percentage); err != nil {
zap.L().Error("Error scanning row", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
heatmap = append(heatmap, item)
}
if err := rows.Err(); err != nil {
zap.L().Error("Error iterating over sample rows", zap.Error(err))
return nil, &model.ApiError{Typ: "ClickHouseError", Err: err}
}
return &heatmap, nil
}