SVLS-8771 fix cloud run jobs shutdown race#47827
Open
apiarian-datadog wants to merge 4 commits intomainfrom
Open
SVLS-8771 fix cloud run jobs shutdown race#47827apiarian-datadog wants to merge 4 commits intomainfrom
apiarian-datadog wants to merge 4 commits intomainfrom
Conversation
Need to make sure that the test actually fails in our ci. Then we can re-enable the fix.
github-actions
bot
added
the
short review
Contributor
Files inventory check summaryFile checks results against ancestor f833cb1c: Results for datadog-agent_7.78.0~devel.git.592.d922361.pipeline.102473987-1_amd64.deb:No change detected |
apiarian-datadog
added
team/serverless-azure-gcp
changelog/no-changelog
Regression DetectorRegression Detector ResultsMetrics dashboard Baseline: b9ff244 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | docker_containers_cpu | % cpu utilization | +0.11 | [-2.91, +3.14] | 1 | Logs |
Fine details of change detection per experiment
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gate_logs | % cpu utilization | +2.71 | [+1.11, +4.30] | 1 | Logs bounds checks dashboard |
| ➖ | docker_containers_memory | memory utilization | +1.65 | [+1.49, +1.81] | 1 | Logs |
| ➖ | otlp_ingest_metrics | memory utilization | +0.78 | [+0.63, +0.94] | 1 | Logs |
| ➖ | tcp_syslog_to_blackhole | ingress throughput | +0.55 | [+0.41, +0.69] | 1 | Logs |
| ➖ | ddot_metrics_sum_cumulative | memory utilization | +0.47 | [+0.33, +0.61] | 1 | Logs |
| ➖ | ddot_logs | memory utilization | +0.46 | [+0.40, +0.53] | 1 | Logs |
| ➖ | quality_gate_idle_all_features | memory utilization | +0.14 | [+0.10, +0.18] | 1 | Logs bounds checks dashboard |
| ➖ | file_to_blackhole_500ms_latency | egress throughput | +0.12 | [-0.27, +0.51] | 1 | Logs |
| ➖ | docker_containers_cpu | % cpu utilization | +0.11 | [-2.91, +3.14] | 1 | Logs |
| ➖ | quality_gate_idle | memory utilization | +0.10 | [+0.05, +0.15] | 1 | Logs bounds checks dashboard |
| ➖ | file_to_blackhole_0ms_latency | egress throughput | +0.08 | [-0.45, +0.61] | 1 | Logs |
| ➖ | file_tree | memory utilization | +0.08 | [+0.02, +0.13] | 1 | Logs |
| ➖ | quality_gate_metrics_logs | memory utilization | +0.02 | [-0.22, +0.26] | 1 | Logs bounds checks dashboard |
| ➖ | tcp_dd_logs_filter_exclude | ingress throughput | +0.01 | [-0.10, +0.12] | 1 | Logs |
| ➖ | file_to_blackhole_100ms_latency | egress throughput | +0.00 | [-0.09, +0.09] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api | ingress throughput | -0.00 | [-0.20, +0.19] | 1 | Logs |
| ➖ | uds_dogstatsd_to_api_v3 | ingress throughput | -0.01 | [-0.20, +0.18] | 1 | Logs |
| ➖ | file_to_blackhole_1000ms_latency | egress throughput | -0.02 | [-0.44, +0.41] | 1 | Logs |
| ➖ | ddot_metrics_sum_delta | memory utilization | -0.06 | [-0.22, +0.10] | 1 | Logs |
| ➖ | otlp_ingest_logs | memory utilization | -0.11 | [-0.20, -0.01] | 1 | Logs |
| ➖ | uds_dogstatsd_20mb_12k_contexts_20_senders | memory utilization | -0.29 | [-0.36, -0.23] | 1 | Logs |
| ➖ | ddot_metrics_sum_cumulativetodelta_exporter | memory utilization | -0.33 | [-0.56, -0.11] | 1 | Logs |
| ➖ | ddot_metrics | memory utilization | -0.49 | [-0.65, -0.33] | 1 | Logs |
Bounds Checks: ✅ Passed
| perf | experiment | bounds_check_name | replicates_passed | observed_value | links |
|---|---|---|---|---|---|
| ✅ | docker_containers_cpu | simple_check_run | 10/10 | 696 ≥ 26 | |
| ✅ | docker_containers_memory | memory_usage | 10/10 | 272.24MiB ≤ 370MiB | |
| ✅ | docker_containers_memory | simple_check_run | 10/10 | 679 ≥ 26 | |
| ✅ | file_to_blackhole_0ms_latency | memory_usage | 10/10 | 0.19GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_0ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_1000ms_latency | memory_usage | 10/10 | 0.23GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_1000ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_100ms_latency | memory_usage | 10/10 | 0.19GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_100ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | file_to_blackhole_500ms_latency | memory_usage | 10/10 | 0.21GiB ≤ 1.20GiB | |
| ✅ | file_to_blackhole_500ms_latency | missed_bytes | 10/10 | 0B = 0B | |
| ✅ | quality_gate_idle | intake_connections | 10/10 | 3 = 3 | bounds checks dashboard |
| ✅ | quality_gate_idle | memory_usage | 10/10 | 173.92MiB ≤ 175MiB | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | intake_connections | 10/10 | 3 = 3 | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | memory_usage | 10/10 | 490.29MiB ≤ 550MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | intake_connections | 10/10 | 3 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_logs | memory_usage | 10/10 | 205.03MiB ≤ 220MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | cpu_usage | 10/10 | 351.29 ≤ 2000 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | intake_connections | 10/10 | 4 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | memory_usage | 10/10 | 402.55MiB ≤ 475MiB | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
Explanation
Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%
Performance changes are noted in the perf column of each table:
- ✅ = significantly better comparison variant performance
- ❌ = significantly worse comparison variant performance
- ➖ = no significant change in performance
A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".
For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:
-
Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.
-
Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.
-
Its configuration does not mark it "erratic".
CI Pass/Fail Decision
✅ Passed. All Quality Gates passed.
- quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
Contributor
Static quality checks✅ Please find below the results from static quality gates Successful checksInfo
25 successful checks with minimal change (< 2 KiB)
On-wire sizes (compressed)
|
Contributor
Author
apiarian-datadog
added
the
qa/done
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 82285ffcfa
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Comment on lines
+142
to
+144
| case ms := <-w.samplesChan: | ||
| w.processSamples(ms) | ||
| default: |
There was a problem hiding this comment.
The new flush path drains samplesChan until it is empty, but in the agent demultiplexer AggregateSamples can keep feeding that channel concurrently while the flush loop waits synchronously for t.blockChan. Under sustained DogStatsD traffic, this loop may never observe the channel empty, so triggerFlush is never reached; that blocks periodic/manual flushes and can also stall shutdown progress because stopChan is not checked inside drainSamples.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
Summary
- Addressed the review feedback on
pkg/aggregator/time_sampler_worker.goby updatingdrainSamples()to consume exactly the snapshotted buffered count (n := len(samplesChan)) without adefaultbranch, preventing premature exit while keeping drain bounded and non-starvable by concurrent producers. pkg/aggregator/time_sampler_worker.goL137-L150 - Added an in-code rationale clarifying why receiving exactly
nitems cannot block in this worker model (single consumer onsamplesChan). pkg/aggregator/time_sampler_worker.goL144-L148 - Committed the change on branch
workas commit1e3946eb, and created the follow-up PR record viamake_prwith a summary and validation steps.
Testing
⚠️ python - <<'PY' ... urllib.request.urlopen('https://go.dev/ref/spec#Channel_types') ... PY(failed due to network proxy/tunnel restriction:403 Forbidden).- ✅
go test ./pkg/aggregator -tags test -run TestFlushIncludesPendingSamples -count=1 - ✅
go test ./pkg/aggregator -tags test -run TestFlushIncludesPendingSamples -count=10
Contributor
Author
There was a problem hiding this comment.
that solution seems a little too clever. if the channel is drained by something else we could end up getting locked, right? this proposed alternative change is okay now, but may be a surprising error in the future.
github-actions
bot
added
medium review
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
1 participant
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.
What does this PR do?
Drain samples on aggregator flush. This way samples that are submitted right at shutdown correctly get flushed.
Motivation
We noticed some metrics being missed in cloud run jobs at shutdown. The metrics get submitted correctly if we enable debug logging, adding time for the aggregator to fully drain.
Describe how you validated your changes
Unit test. Running the test without the fix in CI to confirm that it actually detects the problem. Also deployed to a real cloud run job which shows the issue without the fix and correctly submits metrics with the fix in place.