CASCL-1724: requeue DPA status write on conflict

[clamoriniere]

What does this PR do?

Makes the Cluster Agent DatadogPodAutoscaler (DPA) controller requeue the reconcile when a status write fails instead of silently dropping the update. updateAutoscalerStatusAndUpsert now returns an autoscaling.ProcessResult and sets Requeue when updatePodAutoscalerStatus returns an error (e.g. an HTTP 409 conflict from a stale cached resourceVersion). The two call sites propagate that result. The spec-write path already requeued on error, so no change was needed there.

Motivation

CASCL-1724. A customer reported a DPA that stopped honoring its configured metric and continuously scaled down, with a 409 conflict logged on the leader every ~5 minutes.

Root cause: the status write reuses the resourceVersion read from the informer cache (ObjectMeta: podAutoscaler.ObjectMeta); when that value is stale, UpdateStatus returns 409. In the normal scaling path the resulting error was returned but the reconcile result stayed NoRequeue, so the key was forgotten and only retried on the next informer event (the ~5-minute resync explains the cadence).

Fix approach (idiomatic controller pattern): on a status-write error, requeue and let the reconcile restart. The conflict is caused by a concurrent successful write to the object, which is itself delivered to the informer as a watch event, so a subsequent reconcile reads the up-to-date object from the cache and the write succeeds. Retries are bounded by the workqueue rate-limiter and Process()'s maxRetry. This avoids adding live API reads inside the write path.

Scope note validated during investigation: the 409 does not by itself cause the reported "CPU-only fallback". On a failed status write the controller still upserts the fresh recommendation into its internal store, and horizontal scaling runs before the status write, so the configured metric is still applied. The fallback is driven by main-recommendation staleness and is tracked separately; this PR fixes the conflict handling only.

Describe how you validated your changes

Added unit tests in controller_test.go driving updateAutoscalerStatusAndUpsert against a fake dynamic client:

• TestUpdateAutoscalerStatusAndUpsertRequeuesOnConflict: a 409 on the status subresource makes the reconcile return Requeue and surface the error, and the fresh recommendation is still persisted to the store.
• TestUpdateAutoscalerStatusAndUpsertNoConflict: a successful status write does not requeue (steady state preserved).

Ran the full pkg/clusteragent/autoscaling/workload package test suite (go test -tags 'test kubeapiserver') — all green — and gofmt.

Additional Notes

• Recovery relies on the informer cache being refreshed by the watch event from the write that caused the conflict; no extra API reads are added. A genuinely stale/frozen cache with no further writes would instead be recovered by the informer's own relist.
• Supersedes the earlier retry.RetryOnConflict + live-read implementation on this branch, per review feedback.

---

PR by Bits - View session in Datadog

Comment @DataDog to request changes

[datadog-datadog-prod-us1]

View session in Datadog

Bits Code status: ✅ Done

CI Auto-fix: Disabled | Enable

Comment @DataDog to request changes

[datadog-prod-us1-5]

I can only run on private repositories.

[github-actions]

AI review by Codex (OpenAI) - workflow run

Patch is correct. The conflict retry paths refresh the live object before retrying, preserve concurrent metadata/status state, remain bounded, and include focused regression coverage. I found no actionable issues introduced by this change.

[datadog-datadog-prod-us1]

🎯 Code Coverage (details)
• Patch Coverage: 100.00%
• Overall Coverage: 52.73% (-0.01%)

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 0649698 | Docs | View more details | Give us feedback!}

[dd-octo-sts]

Files inventory check summary

File checks results against ancestor c2fb6573:

Results for datadog-agent_7.84.0~devel.git.575.0649698.pipeline.134011019-1_amd64.deb:

No change detected

Results for datadog-iot-agent_7.84.0~devel.git.575.0649698.pipeline.134011019-1_amd64.deb:

No change detected

[cit-pr-commenter-54b7da]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Job ID: cddaa135-ea1b-48ea-9641-cfe57748fbd8

Baseline: c2fb657
Comparison: 0649698
Diff

Optimization Goals: ✅ No significant changes detected

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	quality_gate_logs	% cpu utilization	+1.77	[+0.90, +2.64]	1	Logs bounds checks dashboard
➖	quality_gate_metrics_logs	memory utilization	+0.83	[+0.60, +1.06]	1	Logs bounds checks dashboard
➖	quality_gate_security_mean_fs_load	memory utilization	+0.30	[+0.27, +0.34]	1	Logs bounds checks dashboard
➖	quality_gate_security_no_fs_load	memory utilization	+0.23	[+0.14, +0.31]	1	Logs bounds checks dashboard
➖	quality_gate_idle	memory utilization	+0.12	[+0.08, +0.17]	1	Logs bounds checks dashboard
➖	quality_gate_idle_all_features	memory utilization	+0.04	[+0.00, +0.07]	1	Logs bounds checks dashboard
➖	quality_gate_security_idle	memory utilization	-0.15	[-0.20, -0.10]	1	Logs bounds checks dashboard
➖	quality_gate_private_action_runner	memory utilization	-0.18	[-0.31, -0.06]	1	Logs bounds checks dashboard
➖	dsd_uds_10mb_3k_timestamped_contexts_memory	memory utilization	-0.78	[-0.99, -0.57]	1	Logs
➖	dsd_uds_10mb_3k_timestamped_contexts_cpu	% cpu utilization	-1.31	[-1.55, -1.07]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	observed_value	links
✅	quality_gate_idle	intake_connections	10/10	4 = 4	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	173.69MiB ≤ 179MiB	bounds checks dashboard
✅	quality_gate_idle	total_bytes_received	10/10	751.32KiB ≤ 819.20KiB	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	4 = 4	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	512.13MiB ≤ 537MiB	bounds checks dashboard
✅	quality_gate_idle_all_features	total_bytes_received	10/10	1.15MiB ≤ 1.25MiB	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	19 ≤ 40	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	215.60MiB ≤ 220MiB	bounds checks dashboard
✅	quality_gate_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard
✅	quality_gate_logs	total_bytes_received	10/10	263.53MiB ≤ 292MiB	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	385.89 ≤ 2000	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	18 ≤ 40	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	419.87MiB ≤ 455MiB	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	0.94GiB ≤ 1.04GiB	bounds checks dashboard
✅	quality_gate_private_action_runner	memory_usage	10/10	71.82MiB ≤ 75MiB	bounds checks dashboard
✅	quality_gate_security_idle	cpu_usage	10/10	29.19 ≤ 100	bounds checks dashboard
✅	quality_gate_security_idle	memory_usage	10/10	325.37MiB ≤ 355MiB	bounds checks dashboard
✅	quality_gate_security_mean_fs_load	cpu_usage	10/10	73.46 ≤ 200	bounds checks dashboard
✅	quality_gate_security_mean_fs_load	memory_usage	10/10	303.86MiB ≤ 335MiB	bounds checks dashboard
✅	quality_gate_security_no_fs_load	cpu_usage	10/10	22.30 ≤ 100	bounds checks dashboard
✅	quality_gate_security_no_fs_load	memory_usage	10/10	309.33MiB ≤ 345MiB	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:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. 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.

3. Its configuration does not mark it "erratic".

CI Pass/Fail Decision

✅ Passed. All Quality Gates passed.

• quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check total_bytes_received: 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_no_fs_load, 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 cpu_usage: 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 intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check total_bytes_received: 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_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_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 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.
• quality_gate_idle, bounds check total_bytes_received: 10/10 replicas passed. Gate passed.
• quality_gate_private_action_runner, bounds check memory_usage: 10/10 replicas passed. Gate passed.

[dd-octo-sts]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor c2fb657
📊 Static Quality Gates Dashboard
🔗 SQG Job

33 successful checks with minimal change (< 2 KiB)

	Quality gate	Current Size
✅	agent_deb_amd64	761.675 MiB
✅	agent_deb_amd64_fips	713.924 MiB
✅	agent_heroku_amd64	312.967 MiB
✅	agent_msi	648.347 MiB
✅	agent_rpm_amd64	761.658 MiB
✅	agent_rpm_amd64_fips	713.908 MiB
✅	agent_rpm_arm64	737.552 MiB
✅	agent_rpm_arm64_fips	693.146 MiB
✅	agent_suse_amd64	761.658 MiB
✅	agent_suse_amd64_fips	713.908 MiB
✅	agent_suse_arm64	737.552 MiB
✅	agent_suse_arm64_fips	693.146 MiB
✅	docker_agent_amd64	820.691 MiB
✅	docker_agent_arm64	821.620 MiB
✅	docker_agent_jmx_amd64	1011.603 MiB
✅	docker_agent_jmx_arm64	1001.312 MiB
✅	docker_cluster_agent_amd64	211.258 MiB
✅	docker_cluster_agent_arm64	224.398 MiB
✅	docker_cws_instrumentation_amd64	7.439 MiB
✅	docker_cws_instrumentation_arm64	6.877 MiB
✅	docker_dogstatsd_amd64	39.545 MiB
✅	docker_dogstatsd_arm64	37.623 MiB
✅	docker_host_profiler_amd64	306.862 MiB
✅	docker_host_profiler_arm64	318.167 MiB
✅	dogstatsd_deb_amd64	30.290 MiB
✅	dogstatsd_deb_arm64	28.302 MiB
✅	dogstatsd_rpm_amd64	30.290 MiB
✅	dogstatsd_suse_amd64	30.290 MiB
✅	iot_agent_deb_amd64	46.568 MiB
✅	iot_agent_deb_arm64	43.204 MiB
✅	iot_agent_deb_armhf	44.015 MiB
✅	iot_agent_rpm_amd64	46.568 MiB
✅	iot_agent_suse_amd64	46.567 MiB

[github-actions]

AI review by Codex (OpenAI) - workflow run

patch is correct — status-write failures now propagate a requeue result through both reconciliation paths, while preserving existing internal-state upserts and retry limits. Tests cover conflict and success cases.