Hi,
Alertmanager pod is stuck with CrashLoopBackOff in Openshift kubernetes
dualstack environment when the alertmanager service uses the same
port(9094) for TCP and UDP. Same build works fine on Openshift kubernetes
non-dualstack environment.
*Environment:*
Openshift cluster with both IPv4 and IPv6 interface.
Aspenmesh-1.11.8-am4
alertmanager - v0.24.0
*Observation:*
For alertmanager service, when tried with different ports for TCP(9094) and
UDP(9095) the deployment is successful.
*Query:*
1. Is there a need to use different ports for UDP and TCP protocols for
alertmanager service when deployed in a dualstack k8s environment?
2. If yes, how does the configured UDP port affect the traffic flow to
the alertmanager service?
This issue is happening on a fresh installation(not upgrade/rollback)
In side-car logs of alertmanager pod,
warning envoy config gRPC config for xxxxxx rejected: Error
adding/updating listener(s) virtualInbound: error adding listener
'0.0.0.0:15006': filter chain '0.0.0.0_9094' has the same matching
rules defined as '0.0.0.0_9094'
virtualInbound6: error adding listener '[::]:15006': filter chain
'::_9094' has the same matching rules defined as '::_9094'
In istiod logs
info ads ADS: new connection for node:alertmanager-0
info ads CDS: PUSH request for node:alertmanager-0 size:365.9kB
info ads EDS: PUSH request for node:alertmanager-0
resources:140 size:62.8kB empty:0 cached:140/140
info ads LDS: PUSH request for node:alertmanager-0 size:648.6kB
info ads RDS: PUSH request for node:alertmanager-0
resources:23 size:43.8kB
warn ads ADS:LDS: ACK ERROR alertmanager-0 Internal:Error
adding/updating listener(s) virtualInbound: error adding listener
'0.0.0.0:15006': filter chain '0.0.0.0_9094' has the same matching
rules defined as '0.0.0.0_9094'
info ads ADS: x.x.x.x:xxxx alertmanager-0 terminated rpc
error: code = Canceled desc = context canceled
--
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 [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/prometheus-users/CAAB105SFc0DJ6qjfGSUwnjyOkpt_z40cAKP5MKggntmG3c5XsQ%40mail.gmail.com.
