HAProxy crashes when monitoring is disabled
Bug
Description
When attempting to deploy with Istio and AuthService enabled, but Monitoring disabled, HA Proxy deploys and crashes because it cannot find services for alert manager and Prometheus. It should not be looking for these since monitoring.enabled: false is set.
Log messages from authservice/authservice-haproxy-sso pod:
[NOTICE] 082/161823 (1) : haproxy version is 2.2.2
[NOTICE] 082/161823 (1) : path to executable is /usr/local/sbin/haproxy
[ALERT] 082/161823 (1) : parsing [/usr/local/etc/haproxy/haproxy.cfg:79] : 'server alertmanager' : could not resolve address 'monitoring-monitoring-kube-alertmanager.monitoring.svc.cluster.local'.
[ALERT] 082/161823 (1) : parsing [/usr/local/etc/haproxy/haproxy.cfg:84] : 'server prometheus' : could not resolve address 'monitoring-monitoring-kube-prometheus.monitoring.svc.cluster.local'.
[ALERT] 082/161823 (1) : Failed to initialize server(s) addr.BigBang Version
1.2.0
Designs
No child items are currently assigned. Use child items to break down this issue into smaller parts.
Activity
added Big Bang Add-Ons haproxy kindbug priority8 + 1 deleted label
added to epic &66 (closed)
mentioned in merge request !404 (merged)
marked this issue as related to #350 (closed)
added teamcore/security label
added stale label
This was resolved, conditional checks for monitoring enabled. https://repo1.dso.mil/platform-one/big-bang/bigbang/-/blob/master/chart/templates/haproxy/haproxy-authservice.yaml#L3
removed stale label
Please register or sign in to reply