[otel_collector_internal_telemetry]: Wrong collector instances are preselected.

[gregkalapos]

Integration Name

OpenTelemetry Collector Internal Telemetry [packages/otel_collector_internal_telemetry]

Dataset Name

No response

Integration Version

doesn't matter

Agent Version

doesn't matter

Agent Output Type

elasticsearch

Elasticsearch Version

9.4

OS Version and Architecture

doesn't matter

Software/API Version

No response

Error Message

No response

Event Original

No response

What did you do?

Go to any of the dashboards from the integration

What did you see?

The dashboards have a preselected collector instance, which is incorrect:

E.g.: https://github.com/elastic/integrations/blob/main/packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-4c979b50-19eb-411c-8cfd-6ffe37fb0092.json#L22

Those selectedOptions lines need to be removed.

Things to check:

• What will be then preselected and is that ok?
• Any option to be bale to select the 1. one - here we should keep in mind OpAMP mgmt. integration

What did you expect to see?

Either an empty selection or a collector that does exist.

Anything else?

No response

[github-actions]

tl;dr: the wrong collector instances are preselected because several otel_collector_internal_telemetry dashboards ship with hardcoded instance.selectedOptions values (specific UUIDs), instead of a neutral default like "--ALL--".

Recommendation
Set the instance control default to "--ALL--" (or no fixed selection) in all affected dashboards, and keep collector selection fully data-driven from service.instance.id/service.name at runtime.

Findings

Hardcoded instance defaults are present in shipped dashboard JSON:

• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-4c979b50-19eb-411c-8cfd-6ffe37fb0092.json:21-23

  "selectedOptions": [
    "2941acdc-a2d7-48a7-afea-f0148eb8d569 (otelcol-contrib)"
  ]

• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-58ba1893-9b1b-46e7-8a32-0765a5dd120a.json:21-23

  "selectedOptions": [
    "4d67ef17-93ff-4197-8cc9-2539dcbd5b17 (otelcol-contrib)"
  ]

• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-8f77b7f6-0ec0-462b-bb59-0607ffc464e1.json:72-74
• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-919a1e90-10fd-4671-9612-bb2dd2f5530a.json:55-57

The instance variable is actively used to filter panel queries, e.g.:

• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-4c979b50-19eb-411c-8cfd-6ffe37fb0092.json:545

  | GROK ?instance """(?<target_instance>[^(]*) \(.*\)"""
  | WHERE target_instance == service.instance.id
  

So a hardcoded default causes dashboards to open pre-filtered to a specific environment’s collector instance.

Reference implementation already exists in the package with neutral default:

• packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-tail-sampling-processor-otel.json:17-19
  ("selectedOptions": ["--ALL--"] for instance)

Verification

I verified by searching dashboard JSON and checking repository history metadata:

$ grep -n "otelcol-contrib)\"" packages/otel_collector_internal_telemetry/kibana/dashboard/*.json
...-4c979b50-...json:22: "2941acdc-a2d7-48a7-afea-f0148eb8d569 (otelcol-contrib)"
...-58ba1893-...json:22: "4d67ef17-93ff-4197-8cc9-2539dcbd5b17 (otelcol-contrib)"
...-8f77b7f6-...json:73: "4d67ef17-93ff-4197-8cc9-2539dcbd5b17 (otelcol-contrib)"
...-919a1e90-...json:56: "302b0851-2149-4bf2-9719-fdb0c64846ef (otelcol-contrib)"

$ grep -n -- '"--ALL--"' packages/otel_collector_internal_telemetry/kibana/dashboard/otel_collector_internal_telemetry-tail-sampling-processor-otel.json | head -n 5
18:"--ALL--"
35:"--ALL--"

Detailed Action Plan

1. Update instance control defaults in these files to neutral ("--ALL--" or empty selection, matching package convention):
   • .../otel_collector_internal_telemetry-4c979b50-19eb-411c-8cfd-6ffe37fb0092.json (line ~21)
   • .../otel_collector_internal_telemetry-58ba1893-9b1b-46e7-8a32-0765a5dd120a.json (line ~21)
   • .../otel_collector_internal_telemetry-8f77b7f6-0ec0-462b-bb59-0607ffc464e1.json (line ~72)
   • .../otel_collector_internal_telemetry-919a1e90-10fd-4671-9612-bb2dd2f5530a.json (line ~55)
2. Keep existing instance option query logic (service.instance.id + service.name) unchanged; only remove environment-specific defaults.
3. Run package checks for dashboard assets and preview dashboards to confirm initial load is not pinned to a single collector.
4. Add changelog entry under packages/otel_collector_internal_telemetry/changelog.yml as bugfix for wrong preselected collector instance defaults.

Related Items

Type	Link	Relevance
Issue	#19403	Report being triaged
PR	#18046	Added ES
PR	#17526	Earlier dashboard split/rework for this package
PR	#19049	Recent dashboard/query bugfix in same package (context)
File	packages/otel_collector_internal_telemetry/changelog.yml:22-36	Shows dashboard evolution/release points

Note

🔒 Integrity filter blocked 2 items

The following items were blocked because they don't meet the GitHub integrity level.

• #19403 issue_read: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #19403 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
  github:
    min-integrity: approved  # merged | approved | unapproved | none

---

What is this? | From workflow: Issue Triage

Give us feedback! React with 🚀 if perfect, 👍 if helpful, 👎 if not.