[Backport 7.62.x] [NPM-4125] Refactor connection direction to use SYN packets

[agent-platform-auto-pr]

Backport f07df0a from #32572.

---

<!--

• Contributors are encouraged to read our CONTRIBUTING documentation.
• Both Contributor and Reviewer Checklists are available at https://datadoghq.dev/datadog-agent/guidelines/contributing/#pull-requests.
• The pull request:
  • Should only fix one issue or add one feature at a time.
  • Must update the test suite for the relevant functionality.
  • Should pass all status checks before being reviewed or merged.
• Commit titles should be prefixed with general area of pull request's change.
• Please fill the below sections if possible with relevant information or links.
  -->

What does this PR do?

This PR fixes connection direction by tying it to state stored by the TCP processor. Since packet capture cannot determine connection direction, a separate tuple type EbpflessTuple is used to store state which omits Direction. When a connection is established, it determines the Direction using TCP state.

Motivation

Fixes many tests in the tracer test suite, and gives us accurate connection direction info.

Describe how you validated your changes

<!--
Validate your changes before merge, ensuring that:

• Your PR is tested by static / unit / integrations / e2e tests
• Your PR description details which e2e tests cover your changes, if any
• The PR description contains details of how you validated your changes. If you validated changes manually and not through automated tests, add context on why automated tests did not fit your changes validation.

If you want additional validation by a second person, you can ask reviewers to do it. Describe how to set up an environment for manual tests in the PR description. Manual validation is expected to happen on every commit before merge.

Any manual validation step should then map to an automated test. Manual validation should not substitute automation, minus exceptions not supported by test tooling yet.
-->

Existing tracer test suite tests are no longer broken:

DD_REMOTE_CONFIGURATION_ENABLED=false TEST_EBPFLESS_OVERRIDE=true sudo -E go test -tags=linux,linux_bpf,npm,process,test ./pkg/network/tracer -v --run TestTracerSuite/eBPFless

New tests were added to the TCPProcessor test suite: (in particular, TestTCPProcessorConnDirection should pass)

go test -race -tags=linux,linux_bpf,npm,process,test ./pkg/network/tracer/connection/ebpfless

Possible Drawbacks / Trade-offs

At this point ebpflessTracer.conns seems past its expiration date to me. I'm thinking about ways to delete ebpflessTracer.conns, maybe make each processor (TCP vs UDP) responsible for holding onto the ConnectionStats in their maps themselves.

Additional Notes

<!--

• Anything else we should know when reviewing?
• Include benchmarking information here whenever possible.
• Include info about alternatives that were considered and why the proposed
  version was chosen.
  -->

[agent-platform-auto-pr]

Uncompressed package size comparison

Comparison with ancestor 234771bfeff3cdf8a1209e4be7be4f069346b3b1

Diff per package

package	diff	status	size	ancestor	threshold
datadog-agent-x86_64-rpm	2.62MB	⚠️	1202.96MB	1200.34MB	140.00MB
datadog-agent-x86_64-suse	2.62MB	⚠️	1202.96MB	1200.34MB	140.00MB
datadog-agent-aarch64-rpm	2.44MB	⚠️	947.05MB	944.61MB	140.00MB
datadog-agent-amd64-deb	2.25MB	⚠️	1193.30MB	1191.05MB	140.00MB
datadog-agent-arm64-deb	2.07MB	⚠️	937.41MB	935.34MB	140.00MB
datadog-heroku-agent-amd64-deb	0.07MB	⚠️	505.28MB	505.21MB	70.00MB
datadog-iot-agent-x86_64-rpm	0.01MB	⚠️	113.42MB	113.42MB	10.00MB
datadog-iot-agent-x86_64-suse	0.01MB	⚠️	113.42MB	113.42MB	10.00MB
datadog-iot-agent-amd64-deb	0.01MB	⚠️	113.35MB	113.35MB	10.00MB
datadog-iot-agent-arm64-deb	0.01MB	⚠️	108.82MB	108.81MB	10.00MB
datadog-iot-agent-aarch64-rpm	0.00MB	✅	108.89MB	108.88MB	10.00MB
datadog-dogstatsd-amd64-deb	0.00MB	✅	78.57MB	78.57MB	10.00MB
datadog-dogstatsd-arm64-deb	0.00MB	✅	55.77MB	55.77MB	10.00MB
datadog-dogstatsd-x86_64-rpm	-0.00MB	✅	78.65MB	78.65MB	10.00MB
datadog-dogstatsd-x86_64-suse	-0.00MB	✅	78.65MB	78.65MB	10.00MB

Decision

⚠️ Warning

[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 aws.create-vm --pipeline-id=53348737 --os-family=ubuntu

Note: This applies to commit af4f6e0

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: 163362be-8b43-42d7-8484-5c2de1cc9747

Baseline: e8dd54c
Comparison: af4f6e0
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	+2.20	[-1.06, +5.46]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	+0.18	[-0.59, +0.96]	1	Logs
➖	file_to_blackhole_0ms_latency_http2	egress throughput	+0.13	[-0.66, +0.93]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	+0.03	[-0.82, +0.88]	1	Logs
➖	file_to_blackhole_300ms_latency	egress throughput	+0.02	[-0.63, +0.66]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	+0.01	[-0.75, +0.78]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	-0.00	[-0.01, +0.01]	1	Logs
➖	file_to_blackhole_0ms_latency_http1	egress throughput	-0.01	[-0.91, +0.89]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	-0.01	[-0.13, +0.10]	1	Logs
➖	quality_gate_idle	memory utilization	-0.06	[-0.09, -0.02]	1	Logs bounds checks dashboard
➖	file_to_blackhole_1000ms_latency_linear_load	egress throughput	-0.12	[-0.58, +0.34]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	-0.14	[-0.23, -0.06]	1	Logs bounds checks dashboard
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.26	[-1.04, +0.52]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-0.28	[-0.35, -0.21]	1	Logs
➖	file_tree	memory utilization	-0.44	[-0.57, -0.31]	1	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	-0.83	[-1.51, -0.16]	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_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	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_idle_all_features, 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.