Skip to content

The alert notebook says collection stopped for a fleet-level store alert #4756

Closed

Description

@erikdarlingdata

A fleet-level store self-alert with no resolution row and no later firing reads "Unknown (not collected since ...)" in the alert notebook. That text says a collector stopped. But store self-alerts have no collector. The service process that evaluates them also serves this page, so the reason for that text, that the instrument is down, does not apply.

Only fleet-level store alerts reach this text through the collector check. An unresolved alert with no server returns earlier, and so does an alert with a blank metric name. No test passes a fleet-level store alert to this code. Two tests only mention the text.

Where (as of 64cd886)

  • The text: Darling/PerformanceMonitor.Darling.Service/AlertNotebookEndpoint.cs:926-938. The fleet-level flag is set at :85-90.
  • The earlier returns: :1069-1072 (no server) and :921-924 (blank metric).
  • The tests that mention the text: Darling/Darling.Tests/AlertNotebookEndpointTests.cs:565 and :664.

Fix

For a fleet-level store alert with no resolution and no later firing, return "No resolution recorded".

Test that must fail first

A fleet-level store alert with no resolution row reads "No resolution recorded".

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't working

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions