Create portable dashboard for otel JVM metrics

[SylvainJuge]

Create a dedicated "portable dashboard" for OTel Java.

• using current state of semantic conventions which are partially stable: https://opentelemetry.io/docs/specs/semconv/runtime/jvm-metrics/
• otel java agent used for testing: 2.3.0 and 1.33.2, with both we need to opt-in by setting -Dotel.instrumentation.runtime.telemetry.emit-experimental-telemetry=true

The same dashboard is being used for the following values of agent.name:

• opentelemetry/java for the 1.x versions of the otel agent
• opentelemetry/java/opentelemetry-java-instrumentation for the upstream agent
• opentelemetry/java/elastic for the Elastic java agent distribution

As a consequence, the older 1.x opentelemetry agents that are not using the stable jvm metrics conventions won't be supported (their metrics were reported in the process.runtime.jvm.* namespace).

Fixes #174445

Checklist

• fix the top table filtering: it seems the time filtering does not work on this one as more services are listed than expected. (1)
• fix the GC heap zones table (2)
• remove UI elements for Java metrics that are no longer displayed, the "curated UI" parts that have been replaced with portable dashboards. This can be done in a follow-up as those are not visible and thus it's mostly cleaning code.
• the documentation on portable dashboards is not up-to-date as it refers a deprecated ID variable.

For (1) and (2), they are properly filtered and displayed when modifying the dashboard and viewing it in lens, somehow the behavior is different when the dashboards are imported in the APM UI.

Screenshots

Dashboard display/edit

Dashboard integrated in APM UI

[apmmachine]

🤖 GitHub comments

Expand to view the GitHub comments

Just comment with:

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

[SylvainJuge]

I confirm that #186167 fixes the issues that were described in this PR and the dashboard now works as expected.

Thus I'll make this PR "ready for review" so we can merge it.
I'll leave the follow-up tasks to the Kibana team.

[kpatticha]

Thanks for improving the otel JVM metrics.

[elasticmachine]

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

[tonyghiani]

synthtrace change LGTM 👌

[kibana-ci]

💚 Build Succeeded

• Buildkite Build
• Commit: 0c6cdd1
• Kibana Serverless Image: docker.elastic.co/kibana-ci/kibana-serverless:pr-182107-0c6cdd1f10eb
• Observability Deployment

Metrics [docs]

Module Count

Fewer modules leads to a faster build time

id	before	after	diff
apm	1885	1886	+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	3.5MB	3.5MB	+38.5KB

History

• 💛 Build #217340 was flaky ee67cf3

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