metrics_collector_progress [SQL0101]
Edit this pageAs it appears on __system.alerts:
metrics_collector_progress
Metrics collection has no recent completed pass during continuous manager lease coverage. The check runs every minute and requires two unhealthy minutes by default. Completion age defaults to the greater of two minutes and three declared collector poll intervals.
For example, system 1 has continuous manager coverage from 10:00 and its last completed pass is at 10:01. With the default two-minute maximum age, completion becomes overdue at 10:03 and an evaluation at 10:05 can activate the alert. A manager coverage gap restarts pending. SQLStreams cannot report an outage while all its checking machinery is stopped.
Inspect the metrics_collector worker and its failure logs. A pass must finish
every read and measurement write before it refreshes completion. The next check
with recent completion resolves the alert under the same key.
Read it through client.System().Alerts().MetricCollectorProgress().Latest(ctx).
Configure its cadence, maximum age, and pending duration through
SystemConfig.MetricCollectorProgressAlert.
Looking for a different code? Search every thread.