Adaptive (memory-feedback) staged-startup pacing - #54080
Draft
GeorgeHahn wants to merge 2 commits into
Draft
Conversation
The V1 pacing was a fixed timer between coarse stages. That's crude on two axes: it's time-based (wastes time when steps are cheap, too fast when heavy) and order-sensitive (two heavy subsystems in the same stage still stack). Add a shared pacer (pkg/util/stagedstart) with two modes, selected by staged_start.mode: - interval (default, = V1): fixed staged_start.stage_interval between steps. - adaptive: run steps one at a time and advance as soon as the process's own memory settles (growth over settle_window < settle_threshold_bytes after step_min), then reclaim. Because no two heavy inits overlap, startup order stops affecting the memory peak — no manual ordering needed. Bounded worst case: each adaptive step is capped at step_max; on timeout the next step proceeds and a warning is logged, so a process under memory pressure makes progress (and fails loudly) instead of hanging. The settle signal is this process's Go-runtime memory retained from the OS via the stdlib runtime/metrics package — portable across every OS, non-STW, and no new dependency. The sampler is injectable so a truer RSS source can be swapped in. All three staged binaries (core agent sequencer, system-probe module loader, security-agent) now share this pacer. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Measurement scaffolding so the SMP comparison exercises adaptive pacing; revisit the shipping default (interval vs adaptive) before merge. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Contributor
Go Package Import DifferencesBaseline: 8d8cc34
|
Contributor
|
🎯 Code Coverage (details) 🔗 Commit SHA: dfced88 | Docs | Datadog PR Page | Give us feedback! |
Contributor
Files inventory check summaryFile checks results against ancestor 8d8cc340: Results for datadog-agent_7.83.0~devel.git.359.dfced88.pipeline.126675171-1_amd64.deb:No change detected |
Contributor
Static quality checks✅ Please find below the results from static quality gates Successful checksInfo
5 successful checks with minimal change (< 2 KiB)
|
Regression DetectorRegression Detector ResultsMetrics dashboard Baseline: 8d8cc34 Optimization Goals: ✅ No significant changes detected
|
| perf | experiment | goal | Δ mean % | Δ mean % CI | trials | links |
|---|---|---|---|---|---|---|
| ➖ | quality_gate_logs | % cpu utilization | +0.92 | [-0.09, +1.93] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_security_mean_fs_load | memory utilization | +0.46 | [+0.42, +0.51] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_metrics_logs | memory utilization | +0.41 | [+0.16, +0.67] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_idle | memory utilization | -0.23 | [-0.28, -0.17] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_idle_all_features | memory utilization | -0.47 | [-0.55, -0.38] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_security_idle | memory utilization | -0.47 | [-0.52, -0.42] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_private_action_runner | memory utilization | -0.48 | [-0.60, -0.36] | 1 | Logs bounds checks dashboard |
| ➖ | quality_gate_security_no_fs_load | memory utilization | -0.94 | [-1.02, -0.86] | 1 | Logs bounds checks dashboard |
Bounds Checks: ✅ Passed
| perf | experiment | bounds_check_name | replicates_passed | observed_value | links |
|---|---|---|---|---|---|
| ✅ | quality_gate_idle | intake_connections | 10/10 | 3 ≤ 4 | bounds checks dashboard |
| ✅ | quality_gate_idle | memory_usage | 10/10 | 148.02MiB ≤ 154MiB | bounds checks dashboard |
| ✅ | quality_gate_idle | total_bytes_received | 10/10 | 734.63KiB ≤ 819.20KiB | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | intake_connections | 10/10 | 3 ≤ 4 | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | memory_usage | 10/10 | 504.45MiB ≤ 512MiB | bounds checks dashboard |
| ✅ | quality_gate_idle_all_features | total_bytes_received | 10/10 | 1.06MiB ≤ 1.25MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | intake_connections | 10/10 | 5 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_logs | memory_usage | 10/10 | 191.06MiB ≤ 195MiB | bounds checks dashboard |
| ✅ | quality_gate_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
| ✅ | quality_gate_logs | total_bytes_received | 10/10 | 251.58MiB ≤ 292MiB | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | cpu_usage | 10/10 | 358.97 ≤ 2000 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | intake_connections | 10/10 | 5 ≤ 6 | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | memory_usage | 10/10 | 408.42MiB ≤ 430MiB | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | missed_bytes | 10/10 | 0B = 0B | bounds checks dashboard |
| ✅ | quality_gate_metrics_logs | total_bytes_received | 10/10 | 0.90GiB ≤ 1.04GiB | bounds checks dashboard |
| ✅ | quality_gate_private_action_runner | memory_usage | 10/10 | 70.79MiB ≤ 75MiB | bounds checks dashboard |
| ✅ | quality_gate_security_idle | cpu_usage | 10/10 | 32.01 ≤ 100 | bounds checks dashboard |
| ✅ | quality_gate_security_idle | memory_usage | 10/10 | 286.22MiB ≤ 330MiB | bounds checks dashboard |
| ✅ | quality_gate_security_mean_fs_load | cpu_usage | 10/10 | 64.70 ≤ 200 | bounds checks dashboard |
| ✅ | quality_gate_security_mean_fs_load | memory_usage | 10/10 | 283.62MiB ≤ 310MiB | bounds checks dashboard |
| ✅ | quality_gate_security_no_fs_load | cpu_usage | 10/10 | 24.45 ≤ 100 | bounds checks dashboard |
| ✅ | quality_gate_security_no_fs_load | memory_usage | 10/10 | 264.15MiB ≤ 320MiB | 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".
Replicate Execution Details
We run multiple replicates for each experiment/variant. However, we allow replicates to be automatically retried if there are any failures, up to 8 times, at which point the replicate is marked dead and we are unable to run analysis for the entire experiment. We call each of these attempts at running replicates a replicate execution. This section lists all replicate executions that failed due to the target crashing or being oom killed.
Note: In the below tables we bucket failures by experiment, variant, and failure type. For each of these buckets we list out the replicate indexes that failed with an annotation signifying how many times said replicate failed with the given failure mode. In the below example the baseline variant of the experiment named experiment_with_failures had two replicates that failed by oom kills. Replicate 0, which failed 8 executions, and replicate 1 which failed 6 executions, all with the same failure mode.
| Experiment | Variant | Replicates | Failure | Logs | Debug Dashboard |
|---|---|---|---|---|---|
| experiment_with_failures | baseline | 0 (x8) 1 (x6) | Oom killed | Debug Dashboard |
The debug dashboard links will take you to a debugging dashboard specifically designed to investigate replicate execution failures.
❌ Retried Profiling Replicate Execution Failures (ddprof)
Note: Profiling replicas may still be executing. See the debug dashboard for up to date status.
| Experiment | Variant | Replicates | Failure | Debug Dashboard |
|---|---|---|---|---|
| quality_gate_idle_all_features | baseline | 10 | Oom killed | Debug Dashboard |
| quality_gate_metrics_logs | baseline | 10 | Oom killed | Debug Dashboard |
| quality_gate_metrics_logs | comparison | 10 | Oom killed | Debug Dashboard |
| quality_gate_security_idle | comparison | 10 | Crashed (exit code: 134) | Debug Dashboard |
CI Pass/Fail Decision
✅ Passed. All Quality Gates passed.
- quality_gate_idle_all_features, bounds check total_bytes_received: 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_security_mean_fs_load, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
- quality_gate_security_mean_fs_load, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check total_bytes_received: 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 missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
- quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_security_no_fs_load, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_security_no_fs_load, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
- quality_gate_security_idle, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
- quality_gate_security_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_private_action_runner, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_logs, bounds check total_bytes_received: 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.
- quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
- quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
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 problem does this solve?
The staged-startup pacing (parent PR #53334) waits a fixed timer between coarse stages. That's crude on two axes:
What changed
A shared pacer (
pkg/util/stagedstart) with two modes, selected bystaged_start.mode:interval(default, = current behavior): fixedstaged_start.stage_intervalbetween steps.adaptive: run steps one at a time and advance as soon as the process's own memory settles (retained-memory growth oversettle_windowdrops belowsettle_threshold_bytes, afterstep_min), then reclaim. Because no two heavy inits overlap, startup order stops affecting the memory peak — the system handles the distribution, no manual ordering.Bounded worst case: each adaptive step is capped at
step_max; on timeout the next step proceeds and logs a warning — so a process under memory pressure makes progress (and fails loudly) rather than appearing to hang, even if that means a later OOM.Cross-OS by construction: the settle signal is this process's Go-runtime memory retained from the OS, read via the stdlib
runtime/metricspackage — portable across every OS, non-stop-the-world, no new dependency. The sampler is injectable so a truer RSS source can be swapped in later. All three staged binaries (core-agent sequencer, system-probe module loader, security-agent) share the pacer.Validation
New unit tests for the pacer use a fake clock + injected sampler (deterministic, OS-independent) and explicitly cover the bounded worst case (memory that never settles proceeds at
step_maxand warns) andstep_min/cancellation. Existing sequencer tests updated for the pacer. The real interval-vs-adaptive comparison (peak and time-to-ready) will come from an SMP run on this branch.Note
Stacked on #53334 (
hahn/staged-startup); review/merge that first. Default is unchanged (interval);adaptiveis opt-in pending SMP comparison.