pkg/dyninst/decode: Implement decoding of template messages

[grantseltzer]

What does this PR do?

Adds support for decoding template messages captured via dynamic instrumentation.

Motivation

Allowing users to specify templates in Go DI/LD using the expression language. For now just variable referencing would be supported.

Describe how you validated your changes

Integration and unit tests.

Additional Notes

[agent-platform-auto-pr]

Go Package Import Differences

Baseline: 6feee7a
Comparison: b3a9d45

binary	os	arch	change
system-probe	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/dyninst/exprlang
system-probe	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/dyninst/exprlang

[ajwerner]

Flushing what I had typed, we discussed some of this offline.

[ajwerner]

did you just manually add this line? that's not really reasonable because it means that when other folks re-run go generate it'll go away. One thing you can do is write a shell script that go generate invokes that will both run stringer and then modify the files to add the relevant tags.

[grantseltzer]

Yes - I added it to one file and was just following the IDE telling me to do so, I missed that they were generated. I can write a script in another PR if needed but it doesn't seem to be causing any issues to not have this build tag here. I've just removed the change for now.

[ajwerner]

index into what? I remain skeptical of the value of this field (commented on that previously).

[grantseltzer]

Removed.

[ajwerner]

nit: periods on sentences to match the style in the rest of the file

[ajwerner]

As I said last time around, I don't understand why this is a map.

[grantseltzer]

As discussed IRL, we now have an EventKind and EventExpressionIndex. This aligns the segment with an expression within an event.

[ajwerner]

nit: this field should be called EventKind or trigger event or something -- it's not really the kind of the expression and it's not using the type RootExpressionKind.

This is really about which event this expression value should be taken from, right?

[ajwerner]

We may want to know more about what's going on here. What if you switched on the type and for unsupported told us the operation?

[ajwerner]

I don't understand why this is keyed on id

[ajwerner]

what is this?

[grantseltzer]

It maps expression names to their index within the event. It enables us to not create duplicate expressions when processing template segments.

[agent-platform-auto-pr]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor 6feee7a

Successful checks

Info

	Quality gate	Delta	On disk size (MiB)	Delta	On wire size (MiB)
✅	agent_deb_amd64	$${+0.04}$$	$${677.11}$$ < $${702.15}$$	$${+0.06}$$	$${165.3}$$ < $${177.79}$$
✅	agent_deb_amd64_fips	$${+0.04}$$	$${671.62}$$ < $${696.65}$$	$${-0.05}$$	$${164.25}$$ < $${176.53}$$
✅	agent_heroku_amd64	$${0}$$	$${336.53}$$ < $${340.18}$$	$${-0}$$	$${89.81}$$ < $${91.08}$$
✅	agent_msi	$${0}$$	$${1013.4}$$ < $${1015.38}$$	$${+0.01}$$	$${148.21}$$ < $${150.78}$$
✅	agent_rpm_amd64	$${+0.04}$$	$${677.1}$$ < $${702.14}$$	$${+0.04}$$	$${167.46}$$ < $${180.53}$$
✅	agent_rpm_amd64_fips	$${+0.04}$$	$${671.61}$$ < $${696.64}$$	$${-0.01}$$	$${166.08}$$ < $${178.79}$$
✅	agent_rpm_arm64	$${+0.04}$$	$${667.11}$$ < $${686.31}$$	$${+0.03}$$	$${153.45}$$ < $${161.22}$$
✅	agent_rpm_arm64_fips	$${+0.04}$$	$${662.64}$$ < $${681.84}$$	$${-0.02}$$	$${152.55}$$ < $${160.18}$$
✅	agent_suse_amd64	$${+0.04}$$	$${677.1}$$ < $${702.14}$$	$${+0.04}$$	$${167.46}$$ < $${180.53}$$
✅	agent_suse_amd64_fips	$${+0.04}$$	$${671.61}$$ < $${696.64}$$	$${-0.01}$$	$${166.08}$$ < $${178.79}$$
✅	agent_suse_arm64	$${+0.04}$$	$${667.11}$$ < $${686.31}$$	$${+0.03}$$	$${153.45}$$ < $${161.22}$$
✅	agent_suse_arm64_fips	$${+0.04}$$	$${662.64}$$ < $${681.84}$$	$${-0.02}$$	$${152.55}$$ < $${160.18}$$
✅	docker_agent_amd64	$${+0.04}$$	$${747.66}$$ < $${773.59}$$	$${+0.02}$$	$${252.1}$$ < $${266.06}$$
✅	docker_agent_arm64	$${+0.04}$$	$${761.62}$$ < $${781.7}$$	$${+0.02}$$	$${242.69}$$ < $${251.85}$$
✅	docker_agent_jmx_amd64	$${+0.04}$$	$${938.53}$$ < $${964.45}$$	$${+0.02}$$	$${320.74}$$ < $${334.68}$$
✅	docker_agent_jmx_arm64	$${+0.04}$$	$${941.08}$$ < $${961.17}$$	$${+0.02}$$	$${307.31}$$ < $${316.44}$$
✅	docker_cluster_agent_amd64	$${-0}$$	$${213.35}$$ < $${213.74}$$	$${+0}$$	$${72.41}$$ < $${73.14}$$
✅	docker_cluster_agent_arm64	$${+0}$$	$${229.24}$$ < $${229.68}$$	$${+0}$$	$${68.67}$$ < $${69.41}$$
✅	docker_cws_instrumentation_amd64	$${-0}$$	$${7.07}$$ < $${7.12}$$	$${-0}$$	$${2.95}$$ < $${3.29}$$
✅	docker_cws_instrumentation_arm64	$${0}$$	$${6.69}$$ < $${6.92}$$	$${+0}$$	$${2.7}$$ < $${3.07}$$
✅	docker_dogstatsd_amd64	$${0}$$	$${38.44}$$ < $${39.3}$$	$${+0}$$	$${14.84}$$ < $${15.76}$$
✅	docker_dogstatsd_arm64	$${0}$$	$${37.12}$$ < $${37.94}$$	$${-0}$$	$${14.28}$$ < $${14.83}$$
✅	dogstatsd_deb_amd64	$${0}$$	$${29.66}$$ < $${30.53}$$	$${+0}$$	$${7.82}$$ < $${8.75}$$
✅	dogstatsd_deb_arm64	$${0}$$	$${28.25}$$ < $${29.11}$$	$${+0}$$	$${6.77}$$ < $${7.71}$$
✅	dogstatsd_rpm_amd64	$${0}$$	$${29.66}$$ < $${30.53}$$	$${-0}$$	$${7.83}$$ < $${8.76}$$
✅	dogstatsd_suse_amd64	$${0}$$	$${29.66}$$ < $${30.53}$$	$${-0}$$	$${7.83}$$ < $${8.76}$$
✅	iot_agent_deb_amd64	$${0}$$	$${42.32}$$ < $${54.97}$$	$${+0}$$	$${11.05}$$ < $${14.45}$$
✅	iot_agent_deb_arm64	$${0}$$	$${40.11}$$ < $${51.9}$$	$${+0}$$	$${9.53}$$ < $${12.63}$$
✅	iot_agent_deb_armhf	$${0}$$	$${40.03}$$ < $${51.84}$$	$${+0}$$	$${9.64}$$ < $${12.74}$$
✅	iot_agent_rpm_amd64	$${0}$$	$${42.32}$$ < $${54.97}$$	$${-0}$$	$${11.06}$$ < $${14.47}$$
✅	iot_agent_suse_amd64	$${0}$$	$${42.32}$$ < $${54.97}$$	$${-0}$$	$${11.06}$$ < $${14.47}$$

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: b22df35a-6bfd-4370-8ea5-ced6c0d5a4ca

Baseline: 6feee7a
Comparison: b3a9d45
Diff

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Regressions in experiments with settings containing erratic: true are ignored.

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
✅	docker_containers_cpu	% cpu utilization	-12.10	[-13.66, -10.54]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	docker_containers_memory	memory utilization	+2.52	[+2.17, +2.86]	1	Logs
➖	quality_gate_metrics_logs	memory utilization	+0.71	[+0.52, +0.91]	1	Logs bounds checks dashboard
➖	quality_gate_idle_all_features	memory utilization	+0.47	[+0.42, +0.51]	1	Logs bounds checks dashboard
➖	otlp_ingest_metrics	memory utilization	+0.31	[+0.19, +0.43]	1	Logs
➖	ddot_logs	memory utilization	+0.31	[+0.24, +0.37]	1	Logs
➖	otlp_ingest_logs	memory utilization	+0.20	[+0.07, +0.33]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	+0.03	[-0.57, +0.63]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	+0.01	[-0.60, +0.62]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	-0.00	[-0.01, +0.01]	1	Logs
➖	file_tree	memory utilization	-0.00	[-0.06, +0.06]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	-0.01	[-0.61, +0.59]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	-0.03	[-0.27, +0.20]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	-0.09	[-0.69, +0.51]	1	Logs
➖	quality_gate_idle	memory utilization	-0.13	[-0.16, -0.09]	1	Logs bounds checks dashboard
➖	quality_gate_logs	% cpu utilization	-0.23	[-3.03, +2.56]	1	Logs bounds checks dashboard
➖	ddot_metrics_sum_cumulative	memory utilization	-0.26	[-0.38, -0.14]	1	Logs
➖	ddot_metrics_sum_cumulativetodelta_exporter	memory utilization	-0.35	[-0.56, -0.14]	1	Logs
➖	ddot_metrics_sum_delta	memory utilization	-0.36	[-0.52, -0.20]	1	Logs
➖	ddot_metrics	memory utilization	-0.49	[-0.65, -0.32]	1	Logs
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	-0.59	[-0.66, -0.52]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-0.81	[-0.88, -0.73]	1	Logs
✅	docker_containers_cpu	% cpu utilization	-12.10	[-13.66, -10.54]	1	Logs

Bounds Checks: ❌ Failed

perf	experiment	bounds_check_name	replicates_passed	links
✅	docker_containers_cpu	simple_check_run	10/10
❌	docker_containers_memory	memory_usage	9/10
✅	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_1000ms_latency	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_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
✅	quality_gate_metrics_logs	cpu_usage	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	bounds checks dashboard
✅	quality_gate_metrics_logs	lost_bytes	10/10	bounds checks dashboard
✅	quality_gate_metrics_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:

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_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_logs, 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_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check lost_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, 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, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.

[grantseltzer]

.

[grantseltzer]

Transposing you comment here:

Rather than enforcing this structure at the point of unmarshaling, I think it'd be better to do it at the point of validation / in the call to GetSegments(). Concretely: consider either keeping []json.RawMessage. Alternatively, what if we just at this layer define a struct that has all three fields and then in validate we assert that only the correct subset are populated? I feel like that's pretty clean and quite efficient. Then when we iterate we can just cast to the proper implementation pointer for the corresponding subset of fields.

I see the advantage of your suggestion as having a concrete type here, as all the other fields of LogProbeCommon are. However I still like having an interface here. As Piotr described when he gave his opinion, it's more idiomatic to type switch rather than treat fields of a struct like a union. Are there any performance benefits to your suggestion?

[grantseltzer]

As discussed IRL, we now have an EventKind and EventExpressionIndex. This aligns the segment with an expression within an event.

[grantseltzer]

Removed.

[grantseltzer]

Yes - I added it to one file and was just following the IDE telling me to do so, I missed that they were generated. I can write a script in another PR if needed but it doesn't seem to be causing any issues to not have this build tag here. I've just removed the change for now.

[grantseltzer]

It maps expression names to their index within the event. It enables us to not create duplicate expressions when processing template segments.