Observability¶
Public observability assets live under deploy/observability.
Prometheus¶
Use deploy/observability/prometheus/namrbd-community-scrape.json as a sample
Prometheus scrape configuration. Replace the example hostnames with your SBS
service, SBS data, gateway, and iSCSI gateway endpoints.
The companion alert file is
deploy/observability/prometheus/namrbd-community-alerts.json. It is a
YAML-compatible Prometheus rule file encoded as JSON so the public Makefile can
lint it with jq.
Grafana¶
Import deploy/observability/grafana/namrbd-community-overview.json into
Grafana and select a Prometheus datasource that scrapes the NAMRBD SBS
and gateway endpoints.
Metric Catalog¶
deploy/observability/metrics/namrbd-community-metrics-catalog.json lists the
public SBS, gateway, and iSCSI gateway metric names and the endpoint that owns
each one.
Alerts¶
The starter alert set covers:
- SBS service readiness
- SBS service leader presence
- SBS data readiness
- gateway readiness
- iSCSI gateway readiness
- SBS node down state
- degraded replicated volumes
- repair and rebalance backlog
- maintenance transition failures
- retired payload cleanup failures
- low SBS data store available capacity ratio
Tune alert thresholds to match the size and availability target of your deployment before using them for paging.
Validate all public observability assets with:
make observability-assets-check