[clusteragent/autoscaling] Allow configuring local fallback activation threshold #37581

jennchenn · 2025-06-02T21:09:16Z

What does this PR do?

Allow configuring local fallback activation threshold by setting the following in the DPA:

spec:
  owner: Local
  ...
  fallback:
    horizontal:
      enabled: true
      triggers:
        staleRecommendationThresholdSeconds: 600

Also change the default threshold to activate local fallback to 30 minutes

Motivation

Previously, staleRecommendationThresholdSeconds controlled when a metric would be considered stale when calculating the local fallback recommendations, even though the documentation had it intended to control when local fallback would be activated.

Describe how you validated your changes

Modify staleRecommendationThresholdSeconds
Verify that local fallback activation occurs following the newly configured threshold value

Possible Drawbacks / Trade-offs

We no longer give users the ability to control which metrics are considered stale for the purposes of local fallback recommendations (by default we use a timeframe of 60s).

Additional Notes

cit-pr-commenter · 2025-06-02T22:08:41Z

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: cbb41efa-e2f1-4924-b8be-832b16c8771d

Baseline: 7a0a040
Comparison: 8547fd6
Diff

Optimization Goals: ✅ No significant changes detected

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	+0.15	[+0.09, +0.21]	1	Logs
➖	docker_containers_memory	memory utilization	+0.12	[+0.06, +0.19]	1	Logs
➖	otlp_ingest_metrics	memory utilization	+0.11	[-0.06, +0.28]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	+0.09	[-0.48, +0.67]	1	Logs
➖	ddot_logs	memory utilization	+0.04	[-0.10, +0.18]	1	Logs
➖	file_to_blackhole_0ms_latency_http2	egress throughput	+0.04	[-0.58, +0.65]	1	Logs
➖	file_to_blackhole_1000ms_latency_linear_load	egress throughput	+0.03	[-0.20, +0.26]	1	Logs
➖	file_to_blackhole_0ms_latency_http1	egress throughput	+0.02	[-0.53, +0.56]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	+0.01	[-0.60, +0.63]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.01	[-0.27, +0.30]	1	Logs
➖	file_to_blackhole_300ms_latency	egress throughput	+0.01	[-0.59, +0.60]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.01, +0.02]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	-0.01	[-0.08, +0.06]	1	Logs bounds checks dashboard
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.05	[-0.70, +0.60]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	-0.08	[-0.66, +0.51]	1	Logs
➖	otlp_ingest_logs	memory utilization	-0.14	[-0.27, -0.02]	1	Logs
➖	ddot_metrics	memory utilization	-0.19	[-0.31, -0.08]	1	Logs
➖	file_tree	memory utilization	-0.56	[-0.71, -0.41]	1	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	-0.66	[-1.53, +0.20]	1	Logs
➖	quality_gate_idle	memory utilization	-0.73	[-0.78, -0.67]	1	Logs bounds checks dashboard
➖	tcp_syslog_to_blackhole	ingress throughput	-0.88	[-0.93, -0.83]	1	Logs
➖	quality_gate_logs	% cpu utilization	-1.78	[-4.52, +0.96]	1	Logs bounds checks dashboard
➖	docker_containers_cpu	% cpu utilization	-2.31	[-5.31, +0.70]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	links
✅	docker_containers_cpu	simple_check_run	10/10
✅	docker_containers_memory	memory_usage	10/10
✅	docker_containers_memory	simple_check_run	10/10
✅	file_to_blackhole_0ms_latency	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency	memory_usage	10/10
✅	file_to_blackhole_0ms_latency_http1	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency_http1	memory_usage	10/10
✅	file_to_blackhole_0ms_latency_http2	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency_http2	memory_usage	10/10
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10
✅	file_to_blackhole_1000ms_latency_linear_load	memory_usage	10/10
✅	file_to_blackhole_100ms_latency	lost_bytes	10/10
✅	file_to_blackhole_100ms_latency	memory_usage	10/10
✅	file_to_blackhole_300ms_latency	lost_bytes	10/10
✅	file_to_blackhole_300ms_latency	memory_usage	10/10
✅	file_to_blackhole_500ms_latency	lost_bytes	10/10
✅	file_to_blackhole_500ms_latency	memory_usage	10/10
✅	quality_gate_idle	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	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_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 lost_bytes: 10/10 replicas passed. Gate 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.

agent-platform-auto-pr · 2025-06-02T22:51:39Z

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor 7a0a040

Successful checks

Info

	Quality gate	Delta	On disk size (MiB)	Delta	On wire size (MiB)
✅	agent_deb_amd64	$${+0}$$	$${695.39}$$ < $${752.99}$$	$${+0.03}$$	$${175.94}$$ < $${187.44}$$
✅	agent_deb_amd64_fips	$${+0}$$	$${693.61}$$ < $${751.36}$$	$${-0.01}$$	$${175.56}$$ < $${187.06}$$
✅	agent_heroku_amd64	$${0}$$	$${358.19}$$ < $${369.68}$$	$${-0}$$	$${96.44}$$ < $${99.55}$$
✅	agent_msi	$${+0}$$	$${957.7}$$ < $${987.01}$$	$${+0.02}$$	$${146.12}$$ < $${150.72}$$
✅	agent_rpm_amd64	$${+0}$$	$${695.38}$$ < $${752.98}$$	$${+0.02}$$	$${177.45}$$ < $${190.03}$$
✅	agent_rpm_amd64_fips	$${+0}$$	$${693.6}$$ < $${751.35}$$	$${-0.01}$$	$${177.3}$$ < $${189.81}$$
✅	agent_rpm_arm64	$${+0}$$	$${685.4}$$ < $${739.42}$$	$${+0}$$	$${160.83}$$ < $${171.23}$$
✅	agent_rpm_arm64_fips	$${+0}$$	$${683.75}$$ < $${737.91}$$	$${+0.01}$$	$${159.99}$$ < $${170.22}$$
✅	agent_suse_amd64	$${+0}$$	$${695.38}$$ < $${752.98}$$	$${+0.02}$$	$${177.45}$$ < $${190.03}$$
✅	agent_suse_amd64_fips	$${+0}$$	$${693.6}$$ < $${751.35}$$	$${-0.01}$$	$${177.3}$$ < $${189.81}$$
✅	agent_suse_arm64	$${+0}$$	$${685.4}$$ < $${739.42}$$	$${+0}$$	$${160.83}$$ < $${171.23}$$
✅	agent_suse_arm64_fips	$${+0}$$	$${683.75}$$ < $${737.91}$$	$${+0.01}$$	$${159.99}$$ < $${170.22}$$
✅	docker_agent_amd64	$${+0}$$	$${779.18}$$ < $${849.39}$$	$${+0.01}$$	$${268.38}$$ < $${288.34}$$
✅	docker_agent_arm64	$${+0}$$	$${792.65}$$ < $${858.97}$$	$${-0}$$	$${255.78}$$ < $${274.36}$$
✅	docker_agent_jmx_amd64	$${+0}$$	$${970.38}$$ < $${1040.59}$$	$${+0.01}$$	$${337.37}$$ < $${357.3}$$
✅	docker_agent_jmx_arm64	$${+0}$$	$${972.44}$$ < $${1038.77}$$	$${+0}$$	$${320.71}$$ < $${339.32}$$
✅	docker_agent_windows1809	$${+0}$$	$${1485.79}$$ < $${1557.19}$$	$${-0}$$	$${487.65}$$ < $${507.33}$$
✅	docker_agent_windows1809_core	$${-0.24}$$	$${6215.51}$$ < $${6287.46}$$	$${0}$$	$${2048.0}$$ < $${2067.76}$$
✅	docker_agent_windows1809_core_jmx	$${-0.2}$$	$${6337.32}$$ < $${6409.04}$$	$${0}$$	$${2048.0}$$ < $${2067.76}$$
✅	docker_agent_windows1809_jmx	$${-0}$$	$${1607.41}$$ < $${1679.14}$$	$${-0.39}$$	$${529.54}$$ < $${550.27}$$
✅	docker_agent_windows2022	$${+0.3}$$	$${1505.22}$$ < $${1576.85}$$	$${-0}$$	$${500.34}$$ < $${520.76}$$
✅	docker_agent_windows2022_core	$${-0.38}$$	$${6188.68}$$ < $${6260.53}$$	$${0}$$	$${2048.0}$$ < $${2067.76}$$
✅	docker_agent_windows2022_core_jmx	$${-0.41}$$	$${6310.25}$$ < $${6382.17}$$	$${0}$$	$${2048.0}$$ < $${2067.76}$$
✅	docker_agent_windows2022_jmx	$${-0.02}$$	$${1626.8}$$ < $${1698.2}$$	$${+0.02}$$	$${542.65}$$ < $${562.99}$$
✅	docker_cluster_agent_amd64	$${-0}$$	$${259.21}$$ < $${259.73}$$	$${+0}$$	$${102.9}$$ < $${103.68}$$
✅	docker_cluster_agent_arm64	$${-0}$$	$${273.64}$$ < $${274.24}$$	$${-0}$$	$${97.58}$$ < $${98.45}$$
✅	docker_cws_instrumentation_amd64	$${0}$$	$${7.08}$$ < $${7.12}$$	$${+0}$$	$${2.95}$$ < $${3.29}$$
✅	docker_cws_instrumentation_arm64	$${0}$$	$${6.69}$$ < $${6.92}$$	$${+0}$$	$${2.7}$$ < $${3.07}$$
✅	docker_dogstatsd_amd64	$${0}$$	$${38.76}$$ < $${39.57}$$	$${+0}$$	$${14.9}$$ < $${15.76}$$
✅	docker_dogstatsd_arm64	$${-0}$$	$${37.4}$$ < $${38.2}$$	$${+0}$$	$${13.91}$$ < $${14.83}$$
✅	dogstatsd_deb_amd64	$${0}$$	$${30.44}$$ < $${31.52}$$	$${+0}$$	$${7.99}$$ < $${8.97}$$
✅	dogstatsd_deb_arm64	$${0}$$	$${29.01}$$ < $${30.08}$$	$${+0}$$	$${6.94}$$ < $${7.92}$$
✅	dogstatsd_rpm_amd64	$${0}$$	$${30.44}$$ < $${31.52}$$	$${+0}$$	$${8.0}$$ < $${8.98}$$
✅	dogstatsd_suse_amd64	$${0}$$	$${30.44}$$ < $${31.52}$$	$${+0}$$	$${8.0}$$ < $${8.98}$$
✅	iot_agent_deb_amd64	$${0}$$	$${50.31}$$ < $${60.17}$$	$${+0}$$	$${12.8}$$ < $${15.82}$$
✅	iot_agent_deb_arm64	$${0}$$	$${47.78}$$ < $${56.94}$$	$${-0}$$	$${11.11}$$ < $${13.86}$$
✅	iot_agent_deb_armhf	$${0}$$	$${47.35}$$ < $${56.41}$$	$${+0}$$	$${11.17}$$ < $${13.86}$$
✅	iot_agent_rpm_amd64	$${0}$$	$${50.31}$$ < $${60.18}$$	$${-0}$$	$${12.82}$$ < $${15.84}$$
✅	iot_agent_rpm_arm64	$${0}$$	$${47.78}$$ < $${56.94}$$	$${-0}$$	$${11.13}$$ < $${13.76}$$
✅	iot_agent_suse_amd64	$${0}$$	$${50.31}$$ < $${60.18}$$	$${-0}$$	$${12.82}$$ < $${15.84}$$

… fallback activated

vboulineau · 2025-06-12T18:41:00Z

pkg/clusteragent/autoscaling/workload/controller.go

@@ -517,19 +517,25 @@ func getActiveScalingSources(currentTime time.Time, podAutoscalerInternal *model
 	fallbackHorizontalScalingValues := podAutoscalerInternal.FallbackScalingValues().Horizontal

 	// If main scaling values are not stale, use those
-	if mainHorizontalScalingValues != nil && !isTimestampStale(currentTime, mainHorizontalScalingValues.Timestamp) {
+	if mainHorizontalScalingValues != nil && !isTimestampStale(currentTime, mainHorizontalScalingValues.Timestamp, nil) {


Shouldn't that check the staleness compared with the setting? If I want a local fallback after 5mins, this will return true while it should not

jennchenn · 2025-06-12T22:45:25Z

/merge

dd-devflow · 2025-06-12T22:45:28Z

View all feedbacks in Devflow UI.

2025-06-12 22:45:28 UTC ℹ️ Start processing command /merge

2025-06-12 22:45:33 UTC ℹ️ MergeQueue: pull request added to the queue

The expected merge time in main is approximately 58m (p90).

2025-06-12 23:23:22 UTC ℹ️ MergeQueue: This merge request was merged

jennchenn added changelog/no-changelog component/autoscaling labels Jun 2, 2025

github-actions bot added the medium review PR review might take time label Jun 2, 2025

jennchenn added the team/containers label Jun 2, 2025

jennchenn added 2 commits June 10, 2025 17:59

Use DPA local fallback stale recommendation threshold to control when…

e7e5fb1

… fallback activated

Check spec is not nil before accessing fallback policy

b608d3a

jennchenn force-pushed the jenn/CASCL-341_support-local-fallback-threshold-config branch from dfaeacc to b608d3a Compare June 10, 2025 21:59

jennchenn added the qa/done QA done before merge and regressions are covered by tests label Jun 10, 2025

jennchenn marked this pull request as ready for review June 10, 2025 22:02

jennchenn requested review from a team as code owners June 10, 2025 22:02

vboulineau reviewed Jun 12, 2025

View reviewed changes

jennchenn added 2 commits June 12, 2025 15:42

Fix comparison to stale fallback timestamp

f03cf46

fixup! Fix comparison to stale fallback timestamp

8547fd6

vboulineau approved these changes Jun 12, 2025

View reviewed changes

jennchenn mentioned this pull request Jun 12, 2025

[datadogpodautoscalers] Update default local fallback threshold to 30m DataDog/datadog-operator#1977

Merged

2 tasks

dd-mergequeue bot merged commit 4bc8e27 into main Jun 12, 2025
239 checks passed

dd-mergequeue bot deleted the jenn/CASCL-341_support-local-fallback-threshold-config branch June 12, 2025 23:23

github-actions bot added this to the 7.68.0 milestone Jun 12, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[clusteragent/autoscaling] Allow configuring local fallback activation threshold #37581

[clusteragent/autoscaling] Allow configuring local fallback activation threshold #37581

Uh oh!

jennchenn commented Jun 2, 2025 •

edited

Loading

Uh oh!

cit-pr-commenter bot commented Jun 2, 2025 •

edited

Loading

Fine details of change detection per experiment

Bounds Checks: ✅ Passed

Explanation

Uh oh!

agent-platform-auto-pr bot commented Jun 2, 2025 •

edited

Loading

Info

Uh oh!

vboulineau Jun 12, 2025

Uh oh!

jennchenn Jun 12, 2025

Uh oh!

jennchenn commented Jun 12, 2025

Uh oh!

dd-devflow bot commented Jun 12, 2025 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

[clusteragent/autoscaling] Allow configuring local fallback activation threshold #37581

[clusteragent/autoscaling] Allow configuring local fallback activation threshold #37581

Uh oh!

Conversation

jennchenn commented Jun 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do?

Motivation

Describe how you validated your changes

Possible Drawbacks / Trade-offs

Additional Notes

Uh oh!

cit-pr-commenter bot commented Jun 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Regression Detector

Regression Detector Results

Optimization Goals: ✅ No significant changes detected

Fine details of change detection per experiment

Bounds Checks: ✅ Passed

Explanation

CI Pass/Fail Decision

Uh oh!

agent-platform-auto-pr bot commented Jun 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Static quality checks

Info

Uh oh!

vboulineau Jun 12, 2025

Choose a reason for hiding this comment

Uh oh!

jennchenn Jun 12, 2025

Choose a reason for hiding this comment

Uh oh!

jennchenn commented Jun 12, 2025

Uh oh!

dd-devflow bot commented Jun 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

jennchenn commented Jun 2, 2025 •

edited

Loading

cit-pr-commenter bot commented Jun 2, 2025 •

edited

Loading

agent-platform-auto-pr bot commented Jun 2, 2025 •

edited

Loading

dd-devflow bot commented Jun 12, 2025 •

edited

Loading