[NPM] Optimize memory allocation in TCP Failures path

[akarpz]

What does this PR do?

There were a few issues with the existing impl:

• the size of the failures map was uncapped
• failures matched to incoming closed connections were not being immediately removed from the map
• we had an unoptimized allocation pattern for failure stats objects

This PR solves these three issues which should help to resolve increased memory usage on a small number of hosts that were likely seeing large spikes in TCP failures

Motivation

Additional Notes

in the tcp load test we see significantly fewer items on heap for the methods in question:

link: https://dddev.datadoghq.com/profiling/comparison?compare_end_A=1722484004715&compare_end_B=1722484002521&compare_query_A=service%3Asystem-probe%20agent-env%3Aadamk-loadtest-base%20&compare_query_B=service%3Asystem-probe%20agent-env%3Aadamk-loadtest-new%20&compare_start_A=1722483104715&compare_start_B=1722483102521&compareValuesMode=absolute&comparisonViz=table&group_by=directory&my_code=enabled&profile_type=heap-live-samples

It also significantly reduced the incidence of orphaned failed connections as expected:

Possible Drawbacks / Trade-offs

Describe how to test/QA your changes

[pr-commenter]

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=40711490 --os-family=ubuntu

Note: This applies to commit 5f5f642

[pr-commenter]

Regression Detector

Regression Detector Results

Run ID: 7a67b306-de55-4548-a55d-9f74b6004542 Metrics dashboard Target profiles

Baseline: 94c9e78
Comparison: 5f5f642

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

No significant changes in experiment optimization goals

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

There were no significant changes in experiment optimization goals at this confidence level and effect size tolerance.

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	links
➖	pycheck_1000_100byte_tags	% cpu utilization	+4.02	[-0.76, +8.81]	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	+2.81	[-10.08, +15.70]	Logs
➖	idle	memory utilization	+0.24	[+0.21, +0.27]	Logs
➖	otel_to_otel_logs	ingress throughput	+0.03	[-0.79, +0.84]	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.00	[-0.00, +0.00]	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.01, +0.01]	Logs
➖	file_tree	memory utilization	-0.04	[-0.12, +0.03]	Logs
➖	basic_py_check	% cpu utilization	-0.90	[-3.48, +1.68]	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	-1.03	[-1.91, -0.16]	Logs

Explanation

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".

[guyarb]

This PR solves these three issues which should help to resolve increased memory usage on a small number of hosts that were likely seeing large spikes in TCP failures

I'd love to see benchmarks in such a pr, load test and dogfooding to support this

[akarpz]

@guyarb thanks for your review, I added some load test results/profile in the PR description. I modified the tcp-load-test slightly to generate more failures and stress test this particular code path

[derekwbrown]

approved for (1) test file owned by wkit

[akarpz]

/merge

[dd-devflow]

🚂 MergeQueue: pull request added to the queue

The median merge time in main is 23m.

Use /merge -c to cancel this operation!