Prometheus does not scrape the monitoring itself #366

Closed
opened 2026-10-25 08:56:24 -07:00 by rays · 1 comment
Owner

Found reviewing what production's traces and metrics leave out, 2026-10-25. Prometheus scrapes ipx, the directory and node, not Tempo, Loki, Alloy or itself. Tempo counts the failure it had on 2026-10-25 (tempodb_blocklist_poll_errors_total) and Alloy what it ships; neither was looked at. Wanted: scrape jobs for all four, and an alert on Tempo's poll errors.

Found reviewing what production's traces and metrics leave out, 2026-10-25. Prometheus scrapes ipx, the directory and node, not Tempo, Loki, Alloy or itself. Tempo counts the failure it had on 2026-10-25 (tempodb_blocklist_poll_errors_total) and Alloy what it ships; neither was looked at. Wanted: scrape jobs for all four, and an alert on Tempo's poll errors.
rays added the enhancement label 2026-10-25 08:56:24 -07:00
Author
Owner

Fixed: Prometheus scrapes prometheus, tempo, loki and alloy (prometheus.yml in /mnt/fast/appdata/prometheus/etc, backup .pre-monitoring); Alloy's command sets --server.http.listen-addr=0.0.0.0:12345 so it can be reached. All up. The Tempo alert reads tempodb_blocklist_tenant_index_age_seconds and tempodb_blocklist_length. Deployed 2026-10-25.

Fixed: Prometheus scrapes prometheus, tempo, loki and alloy (prometheus.yml in /mnt/fast/appdata/prometheus/etc, backup .pre-monitoring); Alloy's command sets --server.http.listen-addr=0.0.0.0:12345 so it can be reached. All up. The Tempo alert reads tempodb_blocklist_tenant_index_age_seconds and tempodb_blocklist_length. Deployed 2026-10-25.
rays closed this issue 2026-10-25 09:35:49 -07:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: rays/ipx#366