Skip to content

codeintel: improve auto-indexing grafana dashboard

Administrator requested to merge nsc/autoindex-dashboard-improvements into main

Created by: Strum355

This PR adds a first iteration "Summary" section to "Code Intelligence > Autoindexing" that is more useful off the bat than the low-level RED metrics below. It brings in some queue and scheduling metrics from the "Executor" and "Worker" dashboards.

It also improves the auto-indexing scheduler panels by increasing the range-vector window to 10m over the old 5m. This is necessary to reduce the noise given that that particular job only runs every 10m by default. cc @bobheadxi for headsup on the new monitoring generator addition

Image uses dot-com data, so the auto-indexing jobs inserted over 5m is empty given its a new metric.

image

Closes https://github.com/sourcegraph/sourcegraph/issues/40342

Test plan

Tested locally using local data (for the new metric) and dotcom data for existing ones.

Merge request reports