hanahmily opened a new pull request, #1172:
URL: https://github.com/apache/skywalking-banyandb/pull/1172
## Summary
Refactor the lifecycle cycle-level metrics
(`banyandb_lifecycle_cycles_total`,
`banyandb_lifecycle_last_run_timestamp_seconds`,
`banyandb_lifecycle_last_run_success`) to carry labels `(remote_node,
remote_role, remote_tier, group)`, matching the form of the parallel
`banyandb_lifecycle_migration_*` family emitted by the queue/pub lifecycle
publisher. Drop `banyandb_lifecycle_self_identity_resolution_total`; its
regression-detection role moves to the new labeled
`cycles_total{remote_node!=""}` plus the existing receiver-side count of empty
`remote_node` on lifecycle `banyandb_queue_sub_total_finished` series.
## Why
The cycle-level metrics were pod-aggregates with no labels, so dashboards
couldn't filter by group or by the (sender_node, sender_role, sender_tier)
tuple. The new labels cross-reference the same tuple computed by `parseGroup`
via `resolveSelfIdentity` and stamped onto the wire via `SetSelfNode`, so the
cycle-level series and the wire-level stamps share the same source. The label
form mirrors the per-message `banyandb_lifecycle_migration_*` family
(file-sync: `banyand/queue/pub/chunked_sync.go:67-82`; batch-write:
`banyand/queue/pub/batch.go:215, 271, 291-292, 421, 471-472, 488, 511, 520,
532`); the two families describe different sides (sender vs destination) and
share the same label form so dashboard matchers and regexes apply to both.
## Changes
- **banyand/backup/lifecycle/service.go**: drop `selfIdentityResolution`
field and `banyandb_lifecycle_self_identity_resolution_total` counter. Register
`cyclesTotal` / `lastRunTimestamp` / `lastRunSuccess` with `cycleLabels :=
[]string{remote_node, remote_role, remote_tier, group}`. Add
`recordCycleGroup(group, senderNode, senderRole, senderTier)` — Inc
`cyclesTotal` per group with the (group, remote_*) tuple, capture the cycle's
last-seen tuple. Add `recordLastRun` — Set both `lastRunTimestamp` and
`lastRunSuccess` with the cycle's last-seen tuple (or empty labels on the
empty-cycle path), first calling `Delete` on the previously-emitted tuple
(`emittedLastRun{Group,Node,Role,Tier}` tracking + `emittedLastRunSet` bool for
the empty-tuple predecessor edge case) so each cycle fully replaces the
previous cycle's series. `action()` resets the captured tuple to empty strings
at the start so scheduler-driven consecutive cycles don't see stale labels.
- **banyand/backup/lifecycle/steps.go**: `parseGroup` now returns the
`(senderNode, senderRole, senderTier)` tuple alongside `*GroupConfig`; the
`resolutionCounter` parameter is dropped. The tuple is the same one
`parseGroup` feeds to `client.SetSelfNode` to populate the wire-level
`SenderNode/Role/Tier`, so the cycle-level stamps and the wire-level stamps
share the same source.
- **banyand/backup/lifecycle/{steps,metrics}_test.go**: adjust to the new
label set. New tests cover: `TestRecordCycleGroupStampsLabeledMetrics` (asserts
`cyclesTotal` Inc'd, gauges NOT touched by `recordCycleGroup`),
`TestRecordLastRunResetsTupleAtActionStart` (empty-cycle path),
`TestRecordLastRunTwoCycleReplaceStaleSeries` (regression: cycle B must Delete
cycle A's tuple), `TestRecordLastRunEmptyThenNonEmptyCycle` (regression: cycle
B must Delete the all-empty tuple from a preceding empty cycle A).
- **test/cases/lifecycle/lifecycle.go**: regexes at lines 261-268 require
all four labels in the alphabetical order Prometheus emits them (`group`,
`remote_node`, `remote_role`, `remote_tier`).
- **docs/operation/grafana-fodc-workload.json**: add `group=~\"$group\"`
matcher to the two lifecycle panel queries (lines 3494, 3558).
- **CHANGES.md**: 0.11.0 entry describing the relabel and the removal of
`banyandb_lifecycle_self_identity_resolution_total`.
## Breaking change
Any alert/panel that pinned the unlabeld form (e.g.
`banyandb_lifecycle_last_run_success == 1`) must update to the labeled form
(`banyandb_lifecycle_last_run_success{group!=""}`).
## Test plan
- [x] `make license-check`
- [x] `make check-req`
- [x] `make build`
- [x] `make lint`
- [x] `make check`
- [x] `go test ./banyand/backup/lifecycle/... -count=1` (PASS, ~12s,
includes 5 new tests)
- [x] Codex critic round 5: APPROVED
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
