1. Fix link in alter

In alert:

Link to grafana dashboard should be https://grafana.stage.devshift.net/d/TjP_nMWMk/operations-rbac-w-cpu?orgId=1&from=now-24h&to=now&timezone=UTC&var-datasource=PDD8BE47D10408F45&var-Endpoint=$__all&var-Namespace=rbac-stage

for stage and accordingly for prod.

now I can see prod link in alter for stage. Alert: https://redhat-internal.slack.com/archives/C05LRFL650V/p1764066567648579 

2. Reduce panels in grafana:

Keep only:

RBAC Consumer-> Relations Excessive lag events (>0.02 rate diff for 10m)

RBAC Consumer -> Relations: Replication event creation rate minus total consumer successful "relations" processed

and convert :

RBAC Replication Event Count 

and 

Consumer Relations Message 

to display just numbers like Replication events increase and Sink event increase in https://grafana.stage.devshift.net/d/ce3ty1vy1gpvkd/kessel-relations-api-data-sync?orgId=1&var-Datasource=PDD8BE47D10408F45&from=now-30m&to=now&timezone=browser

• use also sum(increase ... 

3. Add panel for Add replication event latency metrics

which were added here - https://github.com/RedHatInsights/insights-rbac/pull/2231

• panels for
  • # Average replication latency over 5 minutes: rate(rbac_replication_event_latency_seconds_sum[5m]) / rate(rbac_replication_event_latency_seconds_count[5m])  (one number in panel)
  • # 95th percentile latency: histogram_quantile(0.95, rate(rbac_replication_event_latency_seconds_bucket[5m]))   - graph
  • # 99th percentile latency by event type: histogram_quantile(0.99, sum(rate(rbac_replication_event_latency_seconds_bucket[5m])) by (le, event_type))  -  graph