Auto multiline v2 improve telemetry

[gh123man]

What does this PR do?

• Remove default stats telemetry in favor of regular telemetry counters.
• add a reason value to the truncated tag to easily differentiate between different types of truncated logs
• add a source value to the multiline tag to easily differentiate the source of the multiline aggregation
• rename config logs_config.tag_auto_multi_line_logs -> logs_config.tag_multi_line_logs to better describe it's new behavior (this config was not publicly documented and tied behind the experimental flag so it's safe to change)

Motivation

Describe how to test/QA your changes

1. in datadog.yaml

logs_config:
    experimental_auto_multi_line_detection: true
    max_message_size_bytes: 20
    tag_multi_line_logs: true
    tag_truncated_logs: true

2. send any logs (longer than 20 bytes)
3. they should have a truncated tag,
4. remove the max_message_size_bytes setting
5. send multiline logs

2024-08-13 17:22:24 INFO Main main Some log line 149
2024-08-13 17:22:24 INFO Main main Some log line 150
2024-08-13 17:22:24 SEVERE Main main Exception occur
java.lang.Exception: boom
	at Main.funcd(Main.java:50)
	at Main.funcc(Main.java:49)
	at Main.funcb(Main.java:48)
	at Main.funca(Main.java:47)
	at Main.main(Main.java:29)
2024-08-13 17:22:24 INFO Main main Some log line 151
2024-08-13 17:22:24 INFO Main main Some log line 152

6. Confirm they have an multiline:auto_multiline tag.
7. spot check the agent telemetry page for the auto_multi_line_aggregator_flush metrics.

Possible Drawbacks / Trade-offs

Additional Notes

[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=45313501 --os-family=ubuntu

Note: This applies to commit d450395

[pr-commenter]

Regression Detector

Regression Detector Results

Run ID: b09dc6cc-3706-4d1f-ba47-25cfab5c5577 Metrics dashboard Target profiles

Baseline: a81a777
Comparison: d450395

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	trials	links
➖	basic_py_check	% cpu utilization	+1.34	[-1.39, +4.06]	1	Logs
➖	idle	memory utilization	+0.70	[+0.65, +0.74]	1	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	+0.42	[-0.32, +1.16]	1	Logs
➖	otel_to_otel_logs	ingress throughput	+0.24	[-0.56, +1.04]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	+0.06	[-0.04, +0.16]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.00	[-0.09, +0.10]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.01, +0.01]	1	Logs
➖	file_tree	memory utilization	-0.20	[-0.29, -0.11]	1	Logs
➖	idle_all_features	memory utilization	-0.29	[-0.36, -0.23]	1	Logs
➖	pycheck_lots_of_tags	% cpu utilization	-0.78	[-3.26, +1.71]	1	Logs

Bounds Checks

perf	experiment	bounds_check_name	replicates_passed
✅	idle	memory_usage	10/10

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".

[GustavoCaso]

👏

[jszwedko]

Approving @DataDog/agent-processing-and-routing owned files.

[gh123man]

/merge

[dd-devflow]

🚂 MergeQueue: pull request added to the queue

The median merge time in main is 27m.

Use /merge -c to cancel this operation!