Conversation
The Runs tab now shows a 'v{n}' chip on each run, resolved server-side from
the run's executed_workflow_hash to its workflow_history version (timestamp-
aware so a reverted content hash maps to the version in effect at run time).
Clicking the chip opens the History tab and highlights that version (via
?version=N, which History already rings/expands/scrolls to).
feat(runs): show the workflow version each run executed, linkable to History
keeperhub_workflow_errors_by_workflow was an all-time cumulative count filtered only on status='error'. At prod scale that query matched every error execution across all orgs and joined them before narrowing to managed orgs, tripping the 8s metrics statement_timeout (Postgres 57014) on most scrapes. The catch returned [], so reset() emptied the gauge and it flapped (present ~9% of scrapes), which (a) starved the managed-client alert of data and (b) combined with the alert's offset-delta guard to read the full cumulative count as a 1h delta -> false-positive pages. Change the gauge to a rolling-1h count: add completed_at >= now() - interval '1 hour' to the query, backed by a new partial index idx_workflow_executions_error_completed_at (status='error') so the lookup is an index range scan that finishes in ms and never times out. Metric name and labels are unchanged; only the value semantics move from all-time cumulative to last-hour count, which matches the alert's rolling-60-minute definition and lets the alert read the gauge directly (the companion infra change drops the offset/guard math). Migration is @requires-db-prep: an operator applies the index CONCURRENTLY out-of-band; the in-file CREATE INDEX IF NOT EXISTS is a no-op on prod and a fast build on dev/PR DBs. Refs TECH-48
The per-workflow managed-client error gauge was never added to METRICS_REFERENCE.md when it was introduced. Document it with the windowed (last-1h, not cumulative) semantics and its role in the Sky/Ajna managed-client alerts. Refs TECH-48
Comment cited a non-existent idx_workflow_executions_status_completed_at; the migration creates idx_workflow_executions_error_completed_at. Align the comment with the actual index name (PR #272 review, item 4). Refs TECH-48
…tric fix(metrics): window per-workflow error gauge to last 1h
suisuss
added
db-prepped-prod
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.