Add namespaced azure container app tags to spans and profiles

[duncanpharvey]

What does this PR do?

Adds namespaced tags to spans and profiles from Azure Container Apps.

• aca.resource.id
• aca.resource.group
• aca.subscription.id
• aca.replica.name

Motivation

Provide profiling backend with Azure Container App metadata to enable proper billing. The names legacy tags are too generic. Adding the namespace aca.* makes it clearer where these spans and profiles originated.

https://datadoghq.atlassian.net/browse/SVLS-5848

Describe how to test/QA your changes

Build serverless-init test image from https://github.com/DataDog/datadog-lambda-extension/tree/main/scripts.

VERSION=1.0.0 ./build_serverless_init.sh
cp -r .layers/. ./scripts/bin/
docker build --build-arg="TARGET_ARCH=amd64" --platform linux/amd64 -f Dockerfile.serverless-init.build -t duncanacrtesting.azurecr.io/serverless-init-testing:1.0.0 .
docker push duncanacrtesting.azurecr.io/serverless-init-testing:1.0.0

Deploy to Azure Container Apps

az containerapp up \
  --name node-container-profiling \
  --resource-group duncan-harvey-rg \
  --ingress external \
  --target-port 8080 \
  --env-vars "DD_API_KEY=<api key>" \
             "DD_SITE=datadoghq.com" \
             "DD_TRACE_ENABLED=true" \
             "DD_LOGS_ENABLED=true" \
             "DD_PROFILING_ENABLED=true" \
             "DD_AZURE_SUBSCRIPTION_ID=<subscription_id>" \
             "DD_AZURE_RESOURCE_GROUP=<resource group>" \
             "DD_SERVICE=node-container-profiling" \
             "DD_ENV=dev" \
             "DD_VERSION=1.0.0" \
             "DD_LOG_LEVEL=debug" \
  --image duncanacrtesting.azurecr.io/azure-container-app-hello-world-node:1.0.0

Application Code

var express = require('express');
var app = express();

app.get('/', async function (req, res) {
  console.log('node testing');
  res.send('Hello Datadog!');
});

app.listen(8080, function () {
  console.log('Example app listening on port 8080!');
});

Dockerfile

FROM node:22
WORKDIR /app
COPY package*.json index.js ./
RUN npm install
EXPOSE 8080

COPY --from=duncanacrtesting.azurecr.io/serverless-init-testing:1.0.0 /datadog-init /app/datadog-init

RUN npm install --prefix /dd_tracer/node dd-trace --save
ENV NODE_PATH="/dd_tracer/node/node_modules"
ENV NODE_OPTIONS="--require dd-trace/init"
ENTRYPOINT ["/app/datadog-init"]

CMD ["node", "index.js"]

Possible Drawbacks / Trade-offs

Additional Notes

Follows #30428

[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=47985920 --os-family=ubuntu

Note: This applies to commit 01dfdba

[cit-pr-commenter]

Regression Detector

Regression Detector Results

Run ID: 153df84c-1a17-4578-bfc5-6c865d2c2187 Metrics dashboard Target profiles

Baseline: 04e6c26
Comparison: 01dfdba

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
➖	file_tree	memory utilization	+2.79	[+2.65, +2.93]	1	Logs
➖	pycheck_lots_of_tags	% cpu utilization	+1.72	[-0.84, +4.28]	1	Logs
➖	idle	memory utilization	+0.92	[+0.87, +0.97]	1	Logs bounds checks dashboard
➖	otel_to_otel_logs	ingress throughput	+0.30	[-0.51, +1.10]	1	Logs
➖	idle_all_features	memory utilization	+0.25	[+0.16, +0.35]	1	Logs bounds checks dashboard
➖	basic_py_check	% cpu utilization	+0.24	[-2.65, +3.13]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	+0.03	[-0.22, +0.27]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.02	[-0.09, +0.14]	1	Logs
➖	file_to_blackhole_300ms_latency	egress throughput	+0.02	[-0.17, +0.20]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	+0.00	[-0.33, +0.34]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.01, +0.01]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.02	[-0.50, +0.46]	1	Logs
➖	quality_gate_idle	memory utilization	-0.49	[-0.54, -0.45]	1	Logs bounds checks dashboard
➖	quality_gate_idle_all_features	memory utilization	-0.51	[-0.61, -0.40]	1	Logs bounds checks dashboard
➖	file_to_blackhole_100ms_latency	egress throughput	-0.53	[-0.76, -0.30]	1	Logs
➖	tcp_syslog_to_blackhole	ingress throughput	-1.72	[-1.82, -1.62]	1	Logs
➖	uds_dogstatsd_to_api_cpu	% cpu utilization	-1.96	[-2.69, -1.24]	1	Logs

Bounds Checks

perf	experiment	bounds_check_name	replicates_passed
✅	file_to_blackhole_0ms_latency	memory_usage	10/10
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10
✅	file_to_blackhole_100ms_latency	memory_usage	10/10
✅	file_to_blackhole_300ms_latency	memory_usage	10/10
✅	file_to_blackhole_500ms_latency	memory_usage	10/10
✅	idle	memory_usage	10/10
✅	idle_all_features	memory_usage	10/10
✅	quality_gate_idle	memory_usage	10/10
✅	quality_gate_idle_all_features	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".

[purple4reina]

If this slice isn't going to change, then I'd recommend placing it as a global.

[duncanpharvey]

Done!

datadog-agent/pkg/trace/api/profiles.go

Lines 33 to 42 in 01dfdba

	var azureContainerAppTags = []string{
	"subscription_id",
	"resource_group",
	"resource_id",
	"replicate_name",
	"aca.subscription.id",
	"aca.resource.group",
	"aca.resource.id",
	"aca.replica.name",
	}

[ajgajg1134]

Do you want all of these new aca tags to be shown in the flame graph UI? If not it might be nice to prefix them with either _dd or just _ (this also causes them to be skipped for some types of obfuscation which can be nice)

[duncanpharvey]

Even though this change was motivated by billing I do think that these are helpful tags to have available for customers to use to filter to spans related to a specific resource/subscription/replica.

[ajgajg1134]

Sounds good, although it may be a bit confusing (and inefficient) to have duplicates here, is there a plan to deprecate and remove the "old" names?

[duncanpharvey]

It's a good point because I agree it is messy having these duplicate tags. I chatted with my team internally and there is not a plan to deprecate them at this time since we want to avoid any breaking changes for customers using the "old" tags. That's not to say we will never deprecate them but there would need to be more reasons to warrant a new major version than just removing these duplicate tags.

[peterg17]

LGTM - thanks for implementing more specific tags!

[duncanpharvey]

/merge

[dd-devflow]

🚂 MergeQueue: pull request added to the queue

The median merge time in main is 22m.

Use /merge -c to cancel this operation!