fix(agent): downgrade per-attempt run mirror retry logs to warn - #3722
Conversation
Every retry_scheduled flush outcome of the hosted run chunk mirror logged at error level, and node-sentry converts error-level framework logs into Sentry captures. With retry backoff capped at a few seconds, one degraded window of the run-event append endpoint emitted a Sentry error every ~5s per active run. The condition is self-healing, so log each retry attempt at warn and escalate to error only once consecutiveFailures reaches 5. Terminal stop/disable paths keep their existing error-level reporting. Fixes VERYFRONT-AGENT-3
|
Warning Review limit reached
Next review available in: 51 minutes Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: Repository UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (4)
Comment |
📦 Client bundle boundary
A server module in a client graph aborts hydration in the browser. New leaks fail CI; known leaks are tracked in |
The retry-log change moved the generated API reference line number and added a test file that reshuffled coverage shard 2. That exposed repository-hardening tests that read repository files through process cwd while sibling parallel tests can temporarily chdir. Anchor those reads to the test module URL and refresh the pinned API reference line. Constraint: CI runs docs generation on Deno 2.7.7/Linux; local newer Deno reported broad line-number drift.\nRejected: Regenerate all API docs locally | produced toolchain-only churn unrelated to the PR.\nConfidence: high\nScope-risk: narrow\nTested: deno test --preload=src/testing/preload.ts --no-check --parallel --allow-all --unstable-worker-options --unstable-net src/security/repository-hardening.test.ts src/testing/cwd.test.ts\nTested: deno test --preload=src/testing/preload.ts --no-check --allow-all --unstable-worker-options --unstable-net src/agent/conversation/run-chunk-mirror.test.ts\nTested: deno task lint:cwd-relative-test-reads\nTested: deno fmt --check src/security/repository-hardening.test.ts src/agent/conversation/run-chunk-mirror.ts src/agent/conversation/run-chunk-mirror.test.ts docs/api-reference/veryfront/agent.md\nTested: docker run --rm -u "501:20" -e DENO_DIR=/tmp/deno-cache -v "/private/tmp/veryfront-open-pr-review.CGPobM/pr-3722":/work -w /work denoland/deno:2.7.7 deno task docs:api-reference:check\nTested: deno task coverage:ci:shard -- --shard=2/8 --coverage-dir=/tmp/pr3722-coverage-shard-2.k9XkY0\nTested: deno check src/security/repository-hardening.test.ts src/agent/conversation/run-chunk-mirror.ts\nTested: git diff --check\nNot-tested: Full CI locally.
1 participant
Fixes VERYFRONT-AGENT-3 (495 events — every one a retry attempt, not a distinct failure).
Root cause
recordHostedChunkMirrorRetryScheduledlogged every flush retry at error level;node-sentryconverts every error-level framework log into a Sentry capture, and retry backoff caps at 5s — so one degraded window of the run-event append endpoint emitted a Sentry error every ~5s per active run. The condition is self-healing (failed batches re-queue and flush on recovery); terminal stop/disable paths report separately.Fix
Per-attempt retry logs are now warn-level, escalating to error only when
consecutiveFailuresreachesHOSTED_CHUNK_MIRROR_RETRY_ERROR_THRESHOLD = 5. Terminal stop/disable paths untouched (still error).Testing (red-green TDD)
New
run-chunk-mirror.test.tsdrives the seam below and above the threshold asserting the instrumentation level for each — fails on pre-fix src where every attempt logs error (adversarial revert-check), passes at HEAD.Reviewer notes (from adversarial verification)