feat(observability): instrument durable checkpoint operations - #78
Conversation
|
Warning Review limit reached
Next review available in: 19 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: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (17)
Comment |
|
@coderabbitai review Review exact replacement source head |
🧩 Analysis chain🏁 Script executed: #!/bin/bash
set -euo pipefail
source_commit='64eea7edb28e391634a5d13c83495d20c6388c81'
base_commit='6b2e19e3503ef3671aa2d2f974df7561ba1e1ca0'
echo '== Exact commits and ancestry =='
git show -s --format='source=%H%nparents=%P%nsubject=%s' "$source_commit"
git show -s --format='base=%H%nparents=%P%nsubject=%s' "$base_commit"
git merge-base --is-ancestor "$base_commit" "$source_commit" && echo 'base_is_ancestor=true'
git rev-list --left-right --count "${base_commit}...${source_commit}"
echo '== Changed files =='
git diff --name-status "$base_commit" "$source_commit"
echo '== Diff stat =='
git diff --stat "$base_commit" "$source_commit"
echo '== Source file outline =='
ast-grep outline pg_llm_batch --items all --type class,function 2>/dev/null || true
echo '== Changed Python symbols and observability references =='
git diff --unified=80 "$base_commit" "$source_commit" -- 'pg_llm_batch/**/*.py' ':!**/tests/**'
echo '== Test and documentation diff =='
git diff --unified=30 "$base_commit" "$source_commit" -- 'tests/**' 'AGENTS.md' 'ARCHITECTURE.md' 'CHANGELOG.md' 'CLAUDE.md' 'docs/**' '.github/**'
echo '== Focused repository references =='
rg -n -i -C 3 \
'checkpoint|opentelemetry|trace\.get_tracer|start_as_current_span|start_span|set_attribute|set_status|record_exception|histogram|counter|monotonic|perf_counter' \
pg_llm_batch tests 2>/dev/null || trueLength of output: 186010 🏁 Script executed: #!/bin/bash
set -euo pipefail
source_commit='64eea7edb28e391634a5d13c83495d20c6388c81'
base_commit='6b2e19e3503ef3671aa2d2f974df7561ba1e1ca0'
echo '== Commit relationship =='
git show -s --format='source %H%nparents %P%nsubject %s' "$source_commit"
git merge-base --is-ancestor "$base_commit" "$source_commit" && echo 'base is an ancestor'
git rev-list --left-right --count "${base_commit}...${source_commit}"
echo '== Changed files =='
git diff --name-status "$base_commit" "$source_commit"
echo '== New checkpoint telemetry implementation =='
git show "${source_commit}:pg_llm_batch/checkpoint_telemetry.py" | cat -n
echo '== Package-root export =='
git show "${source_commit}:pg_llm_batch/__init__.py" | cat -n
echo '== Checkpoint store public method signatures =='
ast-grep outline pg_llm_batch/checkpoint_store.py --items all --type class,function --view expanded
rg -n -A18 -B3 '^(class | def (load|save|load_in_transaction|save_in_transaction)\b)' \
pg_llm_batch/checkpoint_store.py
echo '== Focused wrapper tests =='
for test_file in \
tests/test_checkpoint_telemetry.py \
tests/test_checkpoint_telemetry_clock.py \
tests/test_checkpoint_telemetry_span_status.py \
tests/test_checkpoint_telemetry_exception_boundary.py \
tests/test_checkpoint_telemetry_concurrency.py \
tests/test_checkpoint_telemetry_storage_semantics.py \
tests/test_checkpoint_telemetry_public_api.py
do
echo "---- ${test_file} ----"
cat -n "$test_file"
done
echo '== Packaging dependency and type/lint configuration references =='
rg -n -C 3 'opentelemetry|python_requires|requires-python|mypy|ruff' \
pyproject.toml setup.cfg setup.py requirements*.txt 2>/dev/null || trueLength of output: 80130 🏁 Script executed: #!/bin/bash
set -euo pipefail
echo '== Documented optional-API and failure-isolation contract =='
rg -n -i -C 4 \
'optional|StatusCode|status Error|failure isolation|observer|best-effort|constructor|degrade' \
docs/adr/0008-checkpoint-opentelemetry-observability.md \
docs/checkpoint-observability.md \
docs/doctoring/checkpoint-opentelemetry-observability.md \
AGENTS.md ARCHITECTURE.md CHANGELOG.md CLAUDE.md
echo '== Status resolution call sites and existing regression coverage =='
rg -n -C 5 \
'_resolve_error_status|StatusCode\.ERROR|missing_opentelemetry|optional API' \
pg_llm_batch/checkpoint_telemetry.py \
tests/test_checkpoint_telemetry_span_status.py \
tests/test_checkpoint_telemetry_exception_boundary.py
echo '== Requested staged workflow results =='
gh run view 31285605089 --json databaseId,headSha,status,conclusion,event,name,url 2>&1 || true
gh run view 31285605090 --json databaseId,headSha,status,conclusion,event,name,url 2>&1 || trueLength of output: 27517
CI I found one issue:
The remaining inspected behavior meets the requested boundary: fixed low-cardinality attributes, no protected checkpoint identifiers or exception payloads in package-owned telemetry, explicit Error status without a description, unchanged delegation and transaction ownership, operation-local attribute mappings, and package-root export compatibility.
|
|
Superseded by #92. PR #60 advanced from |
1 participant
Linearized replacement for #61
This Draft replays the exact reviewed checkpoint-observability feature delta onto the exact current predecessor
agent/persistent-result-checkpoint-storeat6b2e19e3503ef3671aa2d2f974df7561ba1e1ca0without force-pushing or choosing conflict sides.RCA and replacement boundary
Predecessor #60 was safely replayed onto the current stack after old #61's earlier reconciliation. The old #61 branch therefore diverged from the new predecessor and GitHub reported direct reconciliation probe #77 non-mergeable. Before replay, the five files modified by the #61 feature and present in both former base
efe2c81f8a9cd8301461fe8ae3ce9d5901652185and current predecessor were verified byte-identical:AGENTS.md,ARCHITECTURE.md,CHANGELOG.md,CLAUDE.md, andpg_llm_batch/__init__.py.Replacement head
64eea7edb28e391634a5d13c83495d20c6388c81is exactly one commit ahead and zero behind exact current predecessor6b2e19e3503ef3671aa2d2f974df7561ba1e1ca0and changes exactly the same 17 checkpoint-observability feature files as the former accepted #61 delta: storage-agnostic low-cardinality OpenTelemetry checkpoint spans/metrics, confidentiality and observer-failure isolation, explicit Error status without descriptions, public API export, deterministic tests, ADR/operator/doctoring documentation, and governing contracts.Old #61 is closed unmerged as superseded. No predecessor-head check, review, or approval transfers.
Current staged evidence
64eea7edb28e391634a5d13c83495d20c6388c81.6b2e19e3503ef3671aa2d2f974df7561ba1e1ca0(PR feat(checkpoint): persist resumable result progress #60 branch).31285605089: completed / success on the exact replacement source head.31285605090: completed / success on the exact replacement source head.Because this is still a stacked Draft, default-branch-only organization workflows and final protected-main gates are not inferred as passing.
Merge boundary
Required order remains
.github#790 -> pg-llm-batch#53 -> #55 -> #56 -> #57 -> #58 -> #59 -> #60 -> this PR -> #79.Keep Draft while the earlier stack remains unintegrated. After every prerequisite safely reaches protected
main, reconcile or retarget this slice onto the actual integrated base and regenerate every required exact-head/exact-base CI, security, dependency, SBOM/provenance, packaging, live-integration, automated-review, branch-protection, repository-policy, and Release Acceptance gate. Merge only with zero unresolved valid findings and a qualifying independent non-author GitHubAPPROVEDreview on the unchanged final head. Queued, pending, failed, cancelled, skipped-required, neutral-required, absent, stale-head/base, predecessor, status-only, author-only, or synthetic-only evidence is never acceptance.