[CWS] Change tags retrieval retry logic

[lebauce]

What does this PR do?

Change the logic that retries fetching the tags of a container
when first resolution fails.

Motivation

When we fail to resolve tags, we push the workload back into a queue
so that we retry later. But:

• we retry fetching tags for only one workload, every 10 seconds
• we don't check that the cgroup is still alive
• there is no limit on the number of retries. So if the tagger, for some reasons, fails to resolve the tags for a container, we'll try to resolve the tags for this container forever, filling the queue and potentially forbidding other workloads to be put into the queue.

Describe how to test/QA your changes

Possible Drawbacks / Trade-offs

Additional Notes

[safchain]

is the || needsTagsResolution(workload) needed here ? already checked line 83 no ?

[lebauce]

Yes, it is. It was part of the previous code already. It seems that fetchTags may return an empty list of tags so we retry.

[agent-platform-auto-pr]

Test changes on VM

Use this command from test-infra-definitions to manually test this PR changes on a VM:

inv create-vm --pipeline-id=50269159 --os-family=ubuntu

Note: This applies to commit 8adbcab

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: f308b79e-cf2c-4504-b480-039e5c3be32b

Baseline: a4a88c6
Comparison: 8adbcab
Diff

Optimization Goals: ❌ Significant changes detected

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	basic_py_check	% cpu utilization	+9.10	[+5.06, +13.15]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
❌	basic_py_check	% cpu utilization	+9.10	[+5.06, +13.15]	1	Logs
➖	otel_to_otel_logs	ingress throughput	+0.59	[-0.10, +1.29]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	+0.17	[+0.04, +0.29]	1	Logs bounds checks dashboard
➖	file_to_blackhole_500ms_latency	egress throughput	+0.06	[-0.71, +0.83]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	+0.03	[-0.76, +0.83]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	+0.03	[-0.69, +0.75]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.00	[-0.10, +0.11]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	-0.00	[-0.02, +0.01]	1	Logs
➖	file_to_blackhole_300ms_latency	egress throughput	-0.01	[-0.64, +0.62]	1	Logs
➖	file_to_blackhole_1000ms_latency_linear_load	egress throughput	-0.17	[-0.63, +0.29]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.32	[-1.10, +0.46]	1	Logs
➖	quality_gate_idle	memory utilization	-0.65	[-0.72, -0.57]	1	Logs bounds checks dashboard
➖	pycheck_lots_of_tags	% cpu utilization	-0.67	[-4.12, +2.79]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-0.75	[-0.80, -0.69]	1	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	-0.83	[-1.55, -0.11]	1	Logs
➖	quality_gate_logs	% cpu utilization	-1.93	[-4.86, +1.01]	1	Logs
➖	file_tree	memory utilization	-2.25	[-2.39, -2.11]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	links
✅	file_to_blackhole_0ms_latency	lost_bytes	10/10
✅	file_to_blackhole_0ms_latency	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	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	bounds checks dashboard
✅	quality_gate_logs	lost_bytes	10/10
✅	quality_gate_logs	memory_usage	10/10

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_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.

[spikat]

If I understand this correctly, it may leads to an infinite loop: if we iterate on the chan and call checkTags, and if a workload can't be resolved it will be pushed back on the same chan .. and will be resolved again in the same tick?

[spikat]

I mean not an infinite loop, but it defeat the purpose of capped retries

[lebauce]

/merge

[dd-devflow]

Devflow running: /merge

View all feedbacks in Devflow UI.

---

2024-12-03 14:28:18 UTC ℹ️ MergeQueue: pull request added to the queue

The median merge time in main is 23m.

---

2024-12-03 14:29:30 UTC ⚠️ MergeQueue: This merge request build was cancelled

This merge request build was cancelled

[lebauce]

/remove

[dd-devflow]

Devflow running: /remove

View all feedbacks in Devflow UI.

---

2024-12-03 14:29:30 UTC ℹ️ Devflow: /remove

[lebauce]

/merge

[dd-devflow]

Devflow running: /merge

View all feedbacks in Devflow UI.

---

2024-12-03 15:23:21 UTC ℹ️ MergeQueue: pull request added to the queue

The median merge time in main is 23m.