Simon Pasquier
Simon Pasquier
Alertmanager doesn't use the netstat program so the CVE doesn't really apply. However the next release of Alertmanager will use a patched busybox image.
+100 to what @roidelapluie said. It would an org-wide decision to support expansion of environment variables in configuration file and it isn't something that we've agreed on so far.
> Looking through the log files, we see that the alertmanager doesn't "learn" the firing alerts from the other member of the cluster, but they are received from prometheus instead...
I'm not sure why you've configured `prometheus-alertmanager-headless` in addition to `prometheus-alertmanager-0.prometheus-alertmanager-headless` and `prometheus-alertmanager-1.prometheus-alertmanager-headless`.
Somehow I've read too quickly your initial report and missed the fact that the "issue" happens only when an Alertmanager instance was recreated... What happens most probably is that Prometheus...
You can increase the `group_wait` interval to a value that is large enough so you're guaranteed to receive both alert (a) and alert (b).
There's something off with your configuration. I'd check the configuration from the Status page in the UI and make sure that `group_wait` is 5m for the route.
@AugerC no problem! thanks for the heads-up :)
can you share the full logs?
It would mean that your instance can't keep up with replicating data with its peers. The `alertmanager_cluster_health_score` metric would tell you about your cluster's health (the lower the better, 0...