chore(sidecar): advance contextual-orchestrator pin to 767e67fb (no implicit model timeout) - #2163
Conversation
…mplicit model timeout) Advance ORCHESTRATOR_PIN_SHA from 414f2297 to protected main@767e67fbc6b881a452761f32abb69b9971b9b03b, the merge of contextual-orchestrator#1053. Under the old pin ModelClient defaulted to timeout=90 and the review gateway inherited it, so NVIDIA NIM google/gemma-4-31b-it attempts in the Noema sidecar timed out at exactly 90 s (15 of 27 attempts in fast-mlsirm#1860 run 34748511702) and the gateway answered 502 provider_connection_error after ~20 min of circuit retries. #1053 makes the model timeout null by default and administrator configured per model. Update the sidecar contract test constant, ADR-0003 (current pin and a 2026-09-13 amendment), and CHANGELOG. tests/test_contextual_orchestrator_* and sidecar contract tests: 225 passed. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
|
Warning Review limit reachedNext included review available in 41 minutes. View limit detailsLimit details: You’ve used the included review currently available. You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Review configuration: ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Advanced Run ID: 📒 Files selected for processing (4)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Infrastructure-exception merge record (2026-09-13)Code/security/contract gates on the exact head all pass: agent-review-runtime-quality, coverage-evidence, coverage-source-tree, Bandit, gitleaks, pip-audit, Semgrep, trivy-fs, scorecard, CodeRabbit, Devin Review; locally 225 sidecar/orchestrator contract tests pass. Failing/pending checks are the pre-fix gateway gating its own fix: |
1 participant
Why
Today's Noema failures in fast-mlsirm (#1860 run 34748511702, #1825 run 34752130895) and in contextual-orchestrator itself (#1053 attempts 1–2) trace to the vendored CO pin
414f2297:ModelClient.__init__defaultstimeout=90and the review gateway constructs its client without a timeout, so every NVIDIA NIMgoogle/gemma-4-31b-itattempt hitsTimeoutErrorat exactly 90 s (sidecar artifact 10315556637: 15 of 27 failed attempts at 90.0 s), the orchestrator cycles circuit open/reset on the same route for ~20 min, and the caller receives502 provider_connection_error. contextual-orchestrator#1053 (merged as767e67fb) removes the implicit deadline: model timeout is null by default and administrator-configured per model (model_timeout_seconds), which is the policy ADR-0003 already states.What
scripts/ci/contextual_orchestrator_review_sidecar.sh:ORCHESTRATOR_PIN_SHA414f2297→767e67fbc6b881a452761f32abb69b9971b9b03b.tests/test_contextual_orchestrator_review_sidecar_contract.py:ORCH_PIN_SHAupdated.docs/adr/0003-...: current pin and a 2026-09-13 amendment recording the root cause.CHANGELOG.md: entry with the evidence above.Verification
tests/test_contextual_orchestrator_review_sidecar_contract.py+tests/test_contextual_orchestrator_*/ sidecar tests: 225 passed locally.bash -non the sidecar script.mainpin); it is not claimed here.🤖 Generated with Claude Code