[APM] feat(plugins-apm): add metrics dashboard for opentelemetry-node services

[david-luna]

Summary

This PR add a specific dashboard for metrics being reported by an opentelemetry agent. As for now the UI is showing a default dashboard that does not conform to the document structure and therefore not showing the real values of the metrics.

The current dasshboard

The new dashboard contains 2 panels:

• memory usage in bytes
• memory utilization in %

CPU metrics are not added yet since there is a bug in @opentelemetry/host-metrics. The fix is pending to be merged open-telemetry/opentelemetry-js-contrib#1785

Checklist

N/A

For maintainers

• This was checked for breaking API changes and was labeled appropriately

[elasticmachine]

Pinging @elastic/obs-ux-infra_services-team (Team:obs-ux-infra_services)

[apmmachine]

🤖 GitHub comments

Expand to view the GitHub comments

Just comment with:

• /oblt-deploy : Deploy a Kibana instance using the Observability test environments.
• /oblt-deploy-serverless : Deploy a serverless Kibana instance using the Observability test environments.
• run elasticsearch-ci/docs : Re-trigger the docs validation. (use unformatted text in the comment!)

[david-luna]

@elasticmachine merge upstream

[kpatticha]

@david-luna thanks for adding this.

CPU metrics are not added yet since there is a bug in @opentelemetry/host-metrics. The fix is pending merged open-telemetry/opentelemetry-js-contrib#1785

The fix is merged, should we also add the CPU metrics ?

[david-luna]

The fix is merged, should we also add the CPU metrics ?

Thanks @kpatticha for reviewing this :)

I'd prefer to merge this and create a new PR to add CPU metrics for 2 reasons:

• the current metrics dashboard for Open Telemetry agents is not working, thus not showing relevant data/graphs for the customers. Merging this PR will add valuable information of their services instrumented with OTel
• CPU metrics are a bit more complex since they come with the labels for state and logical_number so I will have to work with Vega to get some aggregations. I'm not experienced in Vega so it'll take some time.

[kpatticha]

nit: I would use opentelemetry_nodejs (underscore) similar to the filename for consistency.

[david-luna]

done!

[kpatticha]

Thanks for adding this @david-luna . The code LGTM, I couldn't test the UI cause I couldn't find any metrics data on edge-light for opentelemetry-node services.

[kpatticha]

Unless I miss something the metrics tab seems to be broken for other oTel services (if not all) on edge-lite.

FYI, @AlexanderWert are you aware of any work or ticket related to the metrics tab in APM?

[david-luna]

Unless I miss something the metrics tab seems to be broken for other oTel services (if not all) on edge-lite.

you're right. The problem is that documents received from Open Telemetry agents have different shape and the fields do not match for our default metrics dashboard

FYI, @AlexanderWert are you aware of any work or ticket related to the metrics tab in APM?

this is the issue for reference elastic/apm#806

[kpatticha]

Unless I miss something the metrics tab seems to be broken for other oTel services (if not all) on edge-lite.

you're right. The problem is that documents received from Open Telemetry agents have different shape and the fields do not match for our default metrics dashboard

FYI, @AlexanderWert are you aware of any work or ticket related to the metrics tab in APM?

this is the issue for reference elastic/apm#806

thanks

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: b09808f

Metrics [docs]

Module Count

Fewer modules leads to a faster build time

id	before	after	diff
apm	1601	1602	+1

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id	before	after	diff
apm	3.1MB	3.2MB	+10.0KB

History

• 💔 Build #189227 failed 779b42a
• 💚 Build #188456 succeeded 7d4dfe3
• 💚 Build #187597 succeeded 0788bcd
• 💛 Build #187027 was flaky d6533e9

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream