Unverified Commit 717e92d3 authored by Michael Schurter's avatar Michael Schurter Committed by GitHub
Browse files

docs: clarify blocked_evals metrics (#13751)

Related to #13740

- blocked_evals.total_blocked is the number of evals blocked for *any*
  reason
- blocked_evals.total_quota_limit is the number of evals blocked by
  quota limits, but critically: their resources are *not* counted in the
  cpu/memory
No related merge requests found
Showing with 2 additions and 2 deletions
+2 -2
......@@ -258,9 +258,9 @@ those listed in [Key Metrics](#key-metrics) above.
| `nomad.nomad.blocked_evals.memory` | Amount of memory requested by blocked evals | Integer | Gauge | datacenter, host, node_class |
| `nomad.nomad.blocked_evals.job.cpu` | Amount of CPU shares requested by blocked evals of a job | Integer | Gauge | host, job, namespace |
| `nomad.nomad.blocked_evals.job.memory` | Amount of memory requested by blocked evals of a job | Integer | Gauge | host, job, namespace |
| `nomad.nomad.blocked_evals.total_blocked` | Count of evals in the blocked state | Integer | Gauge | host |
| `nomad.nomad.blocked_evals.total_blocked` | Count of evals in the blocked state for any reason (cluster resource exhaustion or quota limtis) | Integer | Gauge | host |
| `nomad.nomad.blocked_evals.total_escaped` | Count of evals that have escaped computed node classes. This indicates a scheduler optimization was skipped and is not usually a source of concern. | Integer | Gauge | host |
| `nomad.nomad.blocked_evals.total_quota_limit` | Count of blocked evals due to quota limits | Integer | Gauge | host |
| `nomad.nomad.blocked_evals.total_quota_limit` | Count of blocked evals due to quota limits (the resources for these jobs are *not* counted in other blocked_evals metrics) | Integer | Gauge | host |
| `nomad.nomad.broker.batch_ready` | Count of batch evals ready to be scheduled | Integer | Gauge | host |
| `nomad.nomad.broker.batch_unacked` | Count of unacknowledged batch evals | Integer | Gauge | host |
| `nomad.nomad.broker.eval_waiting` | Time elapsed with evaluation waiting to be enqueued | Nanoseconds | Gauge | eval_id, job, namespace |
......
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment