- Deprecate legacy alertmanager. Are the new alert improvements compatible with legacy? I don't think they are. More importantly, I don't think they should be. It will be a pain to manage it at both places.
- Improve msteamsv2 experience. I have taken alertmanager's upstream and merged it with our custom implementation. Note the use of `titleLink` field to propagate the url of the alert.
- Delete the private http server needed for alertmanager. It's cleanup as part of 1.
* chore(linter): add more linters and deprecate zap
* chore(linter): add more linters and deprecate zap
* chore(linter): add more linters and deprecate zap
* chore(linter): add more linters and deprecate zap
* feat(ruler): base setup for rules and planned maintenance tables
* feat(ruler): more changes for making ruler org aware
* feat(ruler): fix lint
* feat(ruler): update the edit planned maintenance function
* feat(ruler): local testing edits for planned maintenance
* feat(ruler): abstract store and types from rules pkg
* feat(ruler): abstract store and types from rules pkg
* feat(ruler): abstract out store and add migration
* feat(ruler): frontend changes and review comments
* feat(ruler): add back compareAndSelectConfig
* feat(ruler): changes for alertmanager matchers
* feat(ruler): addressed review comments
* feat(ruler): remove the cascade operations from rules table
* feat(ruler): update the template for alertmanager
* feat(ruler): implement the rule history changes
* feat(ruler): implement the rule history changes
* feat(ruler): implement the rule history changes
---------
Co-authored-by: Vibhu Pandey <vibhupandey28@gmail.com>
### Summary
Integrate the new implementations of the alertmanager along with changes to the ruler. This change can be broadly categoried into 3 parts:
#### Frontend
- The earlier `/api/v1/alerts` api was double encoding the response in json and sending it to the frontend. This PR fixes the json response object.
For instance, we have gone from the response `{
"status": "success",
"data": "{\"status\":\"success\",\"data\":[{\"labels\":{\"alertname\":\"[platform][consumer] consumer is above 100% memory utilization\",\"bu\":\"platform\",\"......
}` to the response `{"status":"success","data":[{"labels":{"alertname":"[Metrics] Pod CP......`
- `msteams` has been changed to `msteamsv2` wherever applicable
#### Ruler
The following changes have been done in the ruler component:
- Removal of the old alertmanager and notifier
- The RuleDB methods `Create`, `Edit` and `Delete` have been made transactional
- Introduction of a new `testPrepareNotifyFunc` for sending test notifications
- Integration with the new alertmanager
#### Alertmanager
Although a huge chunk of the alertmanagers have been merged in previous PRs (the list can be found at https://github.com/SigNoz/platform-pod/issues/404), this PR takes care of changes needed in order to incorporate it with the ruler
- Addition of ruleId based matching
- Support for marshalling the global configuration directly from the upstream alertmanager
- Addition of orgId to the legacy alertmanager
- Support for always adding defaults to both routes and receivers while creating them
- Migration to create the required alertmanager tables
- Migration for msteams to msteamsv2 has been added. We will start using msteamv2 config for the new alertmanager and keep using msteams for the old one.
#### Related Issues / PR's
Closes https://github.com/SigNoz/platform-pod/issues/404
Closes https://github.com/SigNoz/platform-pod/issues/176
