We're using a third-party cloud observability platform based on a standard Prometheus + AM architecture to provide metric observability for our on-premise systems. The platform provides standard HTTP APIs to ingest, read and monitor metrics, and exposes configuration for routing alerts in AM.
A key requirement is to integrate alerting with proprietary *escalation* systems that we deploy on-premise to manage alerts (e.g. Incident Management systems). Essentially, our goal to relay the alerts from the cloud platform to our self-hosted Alertmanager that we fully own and integrates well with our escalation systems through various integrations. I was wondering if there is a recommended pattern for relaying alerts from the cloud AM to our own AM on-premise? An experimental approach which seemed feasible is by setting a low *repeat_interval *in cloud AM to send "raw" notifications to a web-hook proxy, which translates a 'notification' back to an 'alert' before sending it to on-premise AM over a secure connection. The on-premise AM on receiving the alerts will apply AM configuration and escalate downstream. I was wondering if there are any issues or gotchas in AM with this approach that we may be overlooking? Is there a better approach to relay alerts between AMs? Samit -- You received this message because you are subscribed to the Google Groups "Prometheus Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to prometheus-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/f3c64015-2e6c-4051-a7cd-bc278067fafcn%40googlegroups.com.
