Open mustdiechik opened 12 months ago
An interesting behavior: on different clusters, it takes different times for the transition to the Running state. From 40 second to up to 4 minutes. I had similar issue with rabbitMQ release in minikube. It looks like the application can not give the load on the processor. The problem was resolved after I changed cpu governor to "performance" mode. Maybe there is a similar problem here.
test environment:
I patched the chart and add additional loadBalancer for calls:
The same configuration with only different loadBalancerIP tested on 4 k8s clusters. One of them - minikube, 3 other - managed by cloud provider (test, stage, prod).
I have strange problem with prod cluster - in 3 other all works as expected. Pod didn`t switch in Running state. 7.10.3 logs:
After changed tag to 7.10.4 pod was up, but after delete\apply release it not swiching in Runnig state. Logs:
If i disable calls plugin - it woks fine.