[CONTP-1437] Migrate CSI driver check to be a core check

[adel121]

What does this PR do?

Migrates the datadog_csi_driver check from a Python OpenMetrics integration (in integrations-core) to a native Go core check, and adds a COAT (Cross-Org Agent Telemetry) profile so that CSI driver metrics are collected as internal agent telemetry.

Changes:

• New Go core check (pkg/collector/corechecks/containers/csi_driver/): Scrapes the CSI driver's Prometheus endpoint (/metrics) and submits datadog.csi_driver.node_publish_volume_attempts.count and datadog.csi_driver.node_unpublish_volume_attempts.count as MonotonicCount metrics. Handles both with and without the _total suffix that Prometheus client libraries may append to counter names.
• COAT integration: Each scraped metric is also fed into comp/core/telemetry counters, enabling Datadog to observe CSI driver adoption and health across all agent deployments without requiring customers to configure anything.
• COAT profile (comp/core/agenttelemetry/impl/config.go): A csi-driver profile is added to defaultProfiles, collecting node_publish_volume_attempts (aggregated by status, type) and node_unpublish_volume_attempts (aggregated by status). Zero-valued metrics are excluded. Collection starts 60s after agent boot and repeats every 15 minutes.
• Registration: The check factory is registered in pkg/commonchecks/corechecks.go and the config directory is added to AGENT_CORECHECKS in tasks/agent.py so conf.yaml.default ships with the agent.
• Autodiscovery config (cmd/agent/dist/conf.d/datadog_csi_driver.d/conf.yaml.default): Discovers the CSI driver container via the csi-driver AD identifier and targets http://%%host%%:5000/metrics.

Motivation

The Python datadog_csi_driver integration in integrations-core collects metrics over OpenMetrics but cannot participate in COAT because COAT only collects from the agent's internal telemetry registry (comp/core/telemetry). Python checks run in the rtloader and have no access to register Go-side telemetry counters.

Migrating to a Go core check enables:

1. COAT visibility: Datadog can observe CSI driver usage (volume mount/unmount success and failure rates) across all customer deployments, without requiring customers to opt in.
2. Reduced overhead: No Python interpreter involvement; the check is a simple HTTP GET + text parse.
3. Simpler dependency chain: The check ships with the agent binary itself rather than as a separately versioned wheel.

Backwards compatibility

This migration is backwards compatible and can ship in the same release that drops the Python integration:

• Same check name (datadog_csi_driver): The Go core check loader takes priority over the Python wheel loader. When both are present, the Go check wins automatically.
• Same metric names: datadog.csi_driver.node_publish_volume_attempts.count and datadog.csi_driver.node_unpublish_volume_attempts.count are identical to the Python integration's output.
• Same autodiscovery identifiers: The csi-driver AD identifier is unchanged, so existing Kubernetes annotations and Helm chart configurations continue to work without modification.
• Same service check: datadog.csi_driver.openmetrics.health reports OK/Critical.
• No customer-facing configuration changes: The openmetrics_endpoint instance config key is preserved.
• Integration is new and internal: The CSI driver integration was recently introduced and is managed by Datadog (deployed via the Datadog Operator / Helm chart), so there is no established external user base relying on Python-specific behavior.

Describe how you validated your changes

Local Kind cluster testing:

1. Deploy agent image with the Go core check and deployed it alongside the Datadog CSI driver in a Kind cluster.
   Sample

datadog:
  apiKeyExistingSecret: datadog-secret
  appKeyExistingSecret: datadog-secret

  csi:
    enabled: true

agents:
  enabled: true

2. Deployed baseline pods with CSI volumes (DSDSocketDirectory type) to trigger NodePublishVolume / NodeUnpublishVolume calls.
   Sample:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: baseline-datadogpy
spec:
  replicas: 3
  selector:
    matchLabels:
      app: baseline-datadogpy
  template:
    metadata:
      labels:
        app: baseline-datadogpy
    spec:
      containers:
        - name: datadogpy
          image: adelhajhassan918/datadogpy-uds-udp
          imagePullPolicy: Always
          volumeMounts:
            - name: datadog-dsd
              mountPath: "/var/run/tmp"
          env:
            - name: USER
              value: "adel"
            - name: DD_DOGSTATSD_URL
              value: unix:///var/run/datadog/dsd/dsd.socket

      volumes:
        - csi:
            driver: k8s.csi.datadoghq.com
            readOnly: true
            volumeAttributes:
              type: DSDSocketDirectory
          name: datadog-dsd

3. Verified the check is running and collecting metrics:

   $ agent status
   datadog_csi_driver
   ------------------
     Instance ID: datadog_csi_driver [OK]
     Total Runs: 34
     Metric Samples: Last Run: 1, Total: 4
     Service Checks: Last Run: 1, Total: 34
   

4. Confirmed the Go check is loaded (not the Python integration) via agent configcheck, verifying configuration source and check loader.
5. Verified metrics appear in the Datadog UI under datadog.csi_driver.node_publish_volume_attempts.count with the expected tags (status, type, path).

COAT validation:

6. Queried the agent telemetry debug endpoint:

   $ curl -sk -H "Authorization: Bearer $(cat /etc/datadog-agent/auth_token)" \
       https://localhost:5001/metadata/agent-telemetry
   

   Confirmed the csi-driver profile payload contains:

   "datadog_csi_driver.node_publish_volume_attempts": {
     "tags": { "status": "success", "type": "DSDSocketDirectory" },
     "type": "counter",
     "value": 12
   },
   "datadog_csi_driver.node_unpublish_volume_attempts": {
     "tags": { "status": "success", "type": "" },
     "type": "counter",
     "value": 9
   }

Additional Notes

• The check handles the Prometheus _total counter suffix normalization defensively. Current CSI driver versions expose counters without _total, but the OpenMetrics spec mandates it and future Prometheus client library upgrades may add it.
• The COAT profile excludes zero-valued metrics (zero_metric: true), so the csi-driver profile only produces payloads after at least one volume operation has occurred.
• Unit tests (check_test.go) cover configuration parsing, successful metric submission, _total suffix handling, and error scenarios (endpoint down, empty response).
• A follow-up PR in integrations-core should deprecate or remove the Python datadog_csi_driver integration.

[agent-platform-auto-pr]

Go Package Import Differences

Baseline: 3f2c64a
Comparison: 376ea45

binary	os	arch	change
agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
agent	windows	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
agent	darwin	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
agent	darwin	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
iot-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
iot-agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver
heroku-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/collector/corechecks/containers/csi_driver

[agent-platform-auto-pr]

Files inventory check summary

File checks results against ancestor 3f2c64a2:

Results for datadog-agent_7.79.0~devel.git.272.376ea45.pipeline.105250456-1_amd64.deb:

Detected file changes:

1 Added files:

• etc/datadog-agent/conf.d/datadog_csi_driver.d/conf.yaml.default (110.0 B)

[agent-platform-auto-pr]

Static quality checks

❌ Please find below the results from static quality gates
Comparison made with ancestor 3f2c64a
📊 Static Quality Gates Dashboard
🔗 SQG Job

Error

	Quality gate	Change	Size (prev → curr → max)
❌	iot_agent_deb_amd64 (on disk)	+20.03 KiB (0.05% increase)	43.289 → 43.309 → 43.290
❌	iot_agent_deb_armhf (on disk)	+20.02 KiB (0.05% increase)	41.084 → 41.103 → 41.100
❌	iot_agent_rpm_amd64 (on disk)	+20.03 KiB (0.05% increase)	43.290 → 43.309 → 43.290
❌	iot_agent_suse_amd64 (on disk)	+20.03 KiB (0.05% increase)	43.290 → 43.309 → 43.290

Gate failure full details

Quality gate	Error type	Error message
iot_agent_deb_amd64	StaticQualityGateFailed	static_quality_gate_iot_agent_deb_amd64 failed! Disk size 43.3 MB exceeds limit of 43.3 MB by 19.3 KB
iot_agent_deb_armhf	StaticQualityGateFailed	static_quality_gate_iot_agent_deb_armhf failed! Disk size 41.1 MB exceeds limit of 41.1 MB by 3.3 KB
iot_agent_rpm_amd64	StaticQualityGateFailed	static_quality_gate_iot_agent_rpm_amd64 failed! Disk size 43.3 MB exceeds limit of 43.3 MB by 19.8 KB
iot_agent_suse_amd64	StaticQualityGateFailed	static_quality_gate_iot_agent_suse_amd64 failed! Disk size 43.3 MB exceeds limit of 43.3 MB by 19.8 KB

Static quality gates prevent the PR to merge!
You can check the static quality gates confluence page for guidance. We also have a toolbox page available to list tools useful to debug the size increase.

Successful checks

Info

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	+27.85 KiB (0.00% increase)	752.713 → 752.740 → 753.380
✅	agent_deb_amd64_fips	+19.85 KiB (0.00% increase)	709.656 → 709.675 → 713.900
✅	agent_heroku_amd64	+20.14 KiB (0.01% increase)	313.318 → 313.338 → 320.580
✅	agent_msi	+23.61 KiB (0.00% increase)	604.874 → 604.897 → 651.440
✅	agent_rpm_amd64	+27.85 KiB (0.00% increase)	752.696 → 752.724 → 753.350
✅	agent_rpm_amd64_fips	+19.85 KiB (0.00% increase)	709.639 → 709.659 → 713.880
✅	agent_rpm_arm64	+20.17 KiB (0.00% increase)	731.115 → 731.135 → 735.290
✅	agent_rpm_arm64_fips	+20.17 KiB (0.00% increase)	691.087 → 691.107 → 696.840
✅	agent_suse_amd64	+27.85 KiB (0.00% increase)	752.696 → 752.724 → 753.350
✅	agent_suse_amd64_fips	+19.85 KiB (0.00% increase)	709.639 → 709.659 → 713.880
✅	agent_suse_arm64	+20.17 KiB (0.00% increase)	731.115 → 731.135 → 735.290
✅	agent_suse_arm64_fips	+20.17 KiB (0.00% increase)	691.087 → 691.107 → 696.840
✅	docker_agent_amd64	+27.85 KiB (0.00% increase)	813.011 → 813.038 → 815.700
✅	docker_agent_arm64	+20.17 KiB (0.00% increase)	816.204 → 816.224 → 821.970
✅	docker_agent_jmx_amd64	+27.87 KiB (0.00% increase)	1003.927 → 1003.954 → 1006.580
✅	docker_agent_jmx_arm64	+20.17 KiB (0.00% increase)	995.898 → 995.918 → 1001.570
✅	iot_agent_deb_arm64	+20.03 KiB (0.05% increase)	40.336 → 40.356 → 40.920

10 successful checks with minimal change (< 2 KiB)

	Quality gate	Current Size
✅	docker_cluster_agent_amd64	203.941 MiB
✅	docker_cluster_agent_arm64	218.419 MiB
✅	docker_cws_instrumentation_amd64	7.142 MiB
✅	docker_cws_instrumentation_arm64	6.689 MiB
✅	docker_dogstatsd_amd64	39.238 MiB
✅	docker_dogstatsd_arm64	37.445 MiB
✅	dogstatsd_deb_amd64	29.881 MiB
✅	dogstatsd_deb_arm64	28.034 MiB
✅	dogstatsd_rpm_amd64	29.881 MiB
✅	dogstatsd_suse_amd64	29.881 MiB

On-wire sizes (compressed)

	Quality gate	Change	Size (prev → curr → max)
❌	iot_agent_deb_amd64	+4.52 KiB (0.04% increase)	11.406 → 11.410 → 12.040
❌	iot_agent_deb_armhf	+5.42 KiB (0.05% increase)	9.940 → 9.946 → 10.620
❌	iot_agent_rpm_amd64	+6.04 KiB (0.05% increase)	11.419 → 11.425 → 12.060
❌	iot_agent_suse_amd64	+6.04 KiB (0.05% increase)	11.419 → 11.425 → 12.060
✅	agent_deb_amd64	neutral	174.788 MiB → 178.360
✅	agent_deb_amd64_fips	-34.24 KiB (0.02% reduction)	165.365 → 165.332 → 172.790
✅	agent_heroku_amd64	+13.05 KiB (0.02% increase)	75.002 → 75.015 → 79.970
✅	agent_msi	+32.0 KiB (0.02% increase)	138.363 → 138.395 → 146.220
✅	agent_rpm_amd64	-22.08 KiB (0.01% reduction)	177.623 → 177.602 → 181.830
✅	agent_rpm_amd64_fips	-30.95 KiB (0.02% reduction)	167.616 → 167.586 → 173.370
✅	agent_rpm_arm64	+11.78 KiB (0.01% increase)	159.529 → 159.540 → 163.060
✅	agent_rpm_arm64_fips	-11.2 KiB (0.01% reduction)	151.396 → 151.385 → 156.170
✅	agent_suse_amd64	-22.08 KiB (0.01% reduction)	177.623 → 177.602 → 181.830
✅	agent_suse_amd64_fips	-30.95 KiB (0.02% reduction)	167.616 → 167.586 → 173.370
✅	agent_suse_arm64	+11.78 KiB (0.01% increase)	159.529 → 159.540 → 163.060
✅	agent_suse_arm64_fips	-11.2 KiB (0.01% reduction)	151.396 → 151.385 → 156.170
✅	docker_agent_amd64	neutral	268.150 MiB → 272.480
✅	docker_agent_arm64	neutral	255.336 MiB → 261.060
✅	docker_agent_jmx_amd64	+2.03 KiB (0.00% increase)	336.799 → 336.801 → 341.100
✅	docker_agent_jmx_arm64	-2.92 KiB (0.00% reduction)	319.977 → 319.974 → 325.620
✅	docker_cluster_agent_amd64	neutral	71.372 MiB → 72.920
✅	docker_cluster_agent_arm64	neutral	67.000 MiB → 68.220
✅	docker_cws_instrumentation_amd64	neutral	2.999 MiB → 3.330
✅	docker_cws_instrumentation_arm64	neutral	2.729 MiB → 3.090
✅	docker_dogstatsd_amd64	neutral	15.175 MiB → 15.820
✅	docker_dogstatsd_arm64	neutral	14.487 MiB → 14.830
✅	dogstatsd_deb_amd64	neutral	7.892 MiB → 8.790
✅	dogstatsd_deb_arm64	neutral	6.780 MiB → 7.710
✅	dogstatsd_rpm_amd64	neutral	7.903 MiB → 8.800
✅	dogstatsd_suse_amd64	neutral	7.903 MiB → 8.800
✅	iot_agent_deb_arm64	+22.26 KiB (0.22% increase)	9.704 → 9.726 → 10.450

[cit-pr-commenter-54b7da]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: 1df4b9fa-a78f-4bc6-8157-c0c8240fb167

Baseline: 3f2c64a
Comparison: 376ea45
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	-1.17	[-4.10, +1.75]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	tcp_syslog_to_blackhole	ingress throughput	+2.35	[+2.20, +2.50]	1	Logs
➖	quality_gate_metrics_logs	memory utilization	+0.40	[+0.16, +0.65]	1	Logs bounds checks dashboard
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	+0.29	[+0.23, +0.36]	1	Logs
➖	ddot_metrics_sum_cumulative	memory utilization	+0.25	[+0.11, +0.40]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	+0.18	[+0.15, +0.22]	1	Logs bounds checks dashboard
➖	ddot_logs	memory utilization	+0.12	[+0.05, +0.19]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	+0.01	[-0.07, +0.09]	1	Logs
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.01	[-0.10, +0.11]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	+0.00	[-0.20, +0.21]	1	Logs
➖	uds_dogstatsd_to_api_v3	ingress throughput	+0.00	[-0.19, +0.19]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	-0.00	[-0.39, +0.39]	1	Logs
➖	otlp_ingest_metrics	memory utilization	-0.02	[-0.18, +0.14]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.02	[-0.46, +0.41]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	-0.08	[-0.61, +0.45]	1	Logs
➖	docker_containers_memory	memory utilization	-0.08	[-0.15, -0.01]	1	Logs
➖	ddot_metrics_sum_delta	memory utilization	-0.11	[-0.28, +0.06]	1	Logs
➖	ddot_metrics	memory utilization	-0.26	[-0.43, -0.08]	1	Logs
➖	quality_gate_idle	memory utilization	-0.26	[-0.31, -0.21]	1	Logs bounds checks dashboard
➖	ddot_metrics_sum_cumulativetodelta_exporter	memory utilization	-0.30	[-0.53, -0.08]	1	Logs
➖	quality_gate_logs	% cpu utilization	-0.33	[-1.88, +1.23]	1	Logs bounds checks dashboard
➖	file_tree	memory utilization	-0.54	[-0.60, -0.49]	1	Logs
➖	otlp_ingest_logs	memory utilization	-1.03	[-1.14, -0.93]	1	Logs
➖	docker_containers_cpu	% cpu utilization	-1.17	[-4.10, +1.75]	1	Logs

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	observed_value	links
✅	docker_containers_cpu	simple_check_run	10/10	691 ≥ 26
✅	docker_containers_memory	memory_usage	10/10	272.23MiB ≤ 370MiB
✅	docker_containers_memory	simple_check_run	10/10	697 ≥ 26
✅	file_to_blackhole_0ms_latency	memory_usage	10/10	0.19GiB ≤ 1.20GiB
✅	file_to_blackhole_0ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10	0.23GiB ≤ 1.20GiB
✅	file_to_blackhole_1000ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_100ms_latency	memory_usage	10/10	0.19GiB ≤ 1.20GiB
✅	file_to_blackhole_100ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_500ms_latency	memory_usage	10/10	0.22GiB ≤ 1.20GiB
✅	file_to_blackhole_500ms_latency	missed_bytes	10/10	0B = 0B
✅	quality_gate_idle	intake_connections	10/10	3 = 3	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	174.64MiB ≤ 175MiB	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	3 = 3	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	490.68MiB ≤ 550MiB	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	3 ≤ 6	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	202.98MiB ≤ 220MiB	bounds checks dashboard
✅	quality_gate_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	355.69 ≤ 2000	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	4 ≤ 6	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	410.08MiB ≤ 475MiB	bounds checks dashboard
✅	quality_gate_metrics_logs	missed_bytes	10/10	0B = 0B	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_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.
• quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate 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_metrics_logs, bounds check memory_usage: 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 intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f091ab8e6a

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[adel121]

@codex

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. More of your lovely PRs please.

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".