mirror of
https://github.com/block/buzz.git
synced 2026-08-18 06:50:31 +02:00
docs: fold onset attribution — 16:40Z organic community creation, not a deploy
Dawn's onset-attribution finding (RESEARCH/BB_PUBLIC_REDIS_ONSET_ATTRIBUTION.md): onset moves 18:00Z -> 16:40Z (15-min bins blurred it); attributed to organic community creation (58 -> 425 in 2h, same 5-min bin as the Redis breakout, flat subs-per-connection ratio excludes a code change). Community creation is now linear (~3,100/day) but Redis load stays exponential — intensification within communities — so the watch metric is channels-per-community. Adds the four-account cacheclusterid Datadog trap (must scope aws_account:433851229429). No planning number changes: 3.45d doubling and 5.2-9.0d band stand. Co-authored-by: npub1qyvc0c5kl4gqv2fd97fsk46tu378sqgy35vc83rvgfwne90sel7s0ed67d <011987e296fd5006292d2f930b574be47c7801048d1983c46c425d3c95f0cffd@buzz.block.builderlab.xyz> Signed-off-by: npub1qyvc0c5kl4gqv2fd97fsk46tu378sqgy35vc83rvgfwne90sel7s0ed67d <011987e296fd5006292d2f930b574be47c7801048d1983c46c425d3c95f0cffd@buzz.block.builderlab.xyz>
This commit is contained in:
parent
de1fc0a1ff
commit
bfb1cc58fe
@@ -90,13 +90,25 @@ profile `bb-public-operations-ro`).
|
||||
speed; r8g (Graviton4) buys ~25–30% single-thread ≈ 1.1–1.3 days at the
|
||||
fitted 3.45-day doubling (1.25–1.30x → Td·log₂(M); vendor range, not
|
||||
benchmarked by anyone in this effort).
|
||||
- **Fit window (measured, P3):** the curve has a sharp onset at 2026-07-21
|
||||
~18:00Z — flat ~0.55% for the preceding five days, then exponential within
|
||||
one 15-min bin (subscribed channels 74→1303 in four hours at a pinned
|
||||
- **Fit window (measured, P3; onset attributed 2026-07-30):** the curve has a
|
||||
sharp onset at 2026-07-21 **16:40Z** (5-min bins; the earlier ~18:00Z was a
|
||||
15-min-bin artifact) — flat ~0.55% for the preceding five days, then
|
||||
exponential (subscribed channels 74→1303 in four hours at a pinned
|
||||
connection count). Growth fits must start after the onset or they mix two
|
||||
regimes. `curr_connections` moves in pool-config plateaus (40.0 → 74.9 →
|
||||
~280) and must not be read as demand. The onset is unattributed; confirming
|
||||
what reached bb-public at 18:00Z needs deploy-history access.
|
||||
~280) and must not be read as demand. **The onset is not a deploy** — it is
|
||||
organic community creation (58 → 425 communities in two hours, first move in
|
||||
the same 5-min bin as the Redis breakout; one pub/sub topic per community
|
||||
makes the chain mechanical, and subscriptions-per-connection stayed flat
|
||||
across the onset, which excludes a subscription-behavior code change).
|
||||
Community creation has since gone **linear** (~3,100/day) while Redis load
|
||||
remains **exponential** — the exponential is carried by intensification
|
||||
within communities, so the curve-break indicator to watch is
|
||||
**channels-per-community** (rising as of Jul 30), not community count.
|
||||
Receipts: `RESEARCH/BB_PUBLIC_REDIS_ONSET_ATTRIBUTION.md`. Measurement
|
||||
trap: `cacheclusterid:buzz-001-001` exists in **four** AWS accounts, so
|
||||
every ElastiCache query must scope `aws_account:433851229429` — a bare-tag
|
||||
`avg:` dilutes engine CPU ~3.4x.
|
||||
|
||||
## System Model
|
||||
|
||||
|
||||
Reference in New Issue
Block a user