Alerting in the Prometheus Ecosystem: The Past, Present and Future
summary
Explains why the Alertmanager is horizontally scalable inside Cortex/Mimir by shard by tenant id, using hash rings for routing, gRPC for notification-state replication, object storage for push/pull state, and a Dynamo-style quorum for writes; this design runs in Grafana Cloud for nine months without major cost or failures. Describes embedding Alertmanager into Grafana to get grouping, silences and time intervals, adds receiver testing and notification history plans, and intends to upstream improvements.
AI-written summary · about
chapters (8)
- 0:00Start
- 1:26Show of hands
- 2:30The beginning
- 3:52Alertmanager Alert
- 8:59Scaling the Alertmanager
- 19:23Feedback
- 19:53Testing Receivers
- 20:36Alert Pipeline Insights
related talks
50mMetrics and alerting with Prometheus and Grafana
29mParanoid Observability –Success Stories and Misfortunes observing an observability stack
18mAlert Analytics
36mUnderstanding and Extending Prometheus AlertManager
20mThe Evolution of Grafana Alerting
33mAlerting in Cloud Native Environments
32mAlertmanager on Its Way to High Availability
31mEvolution of Metric Monitoring and Alerting: Upgrade Your Prometheus Today