Description of problem: Create a cluster with 2 compute nodes, after creating app see where the mysql pod is created. ssh to that node and stop cri-o service. mysql pod never goes to 0/1 Ready state, it stays on 1/1 Ready state. On docker runtime cluster that pod immediately goes to 0/1 ready state. after around 6 mins pod gets recreated on the other Ready node, same time gap noticed for creation on docker runtime cluster also( not sure why the delay is 6 min, could be another bz) Version-Release number of selected component (if applicable): openshift v3.10.0-0.32.0 kubernetes v1.10.0+b81c8f8 etcd 3.2.16 How reproducible: Always on CRIO Steps to Reproduce: 1. create a cluster with crio runtime 2. create cakephp-mysql app using template 3. ssh to the node where mysql pod is running and stop crio service Actual results: No change in pod status for around 6 mins, it shows 1/1 Ready Expected results: Pod status should change immediately to 0/1 Ready. Docker runtime behaves correct. Additional info:
Created another bz with a better explanation. https://bugzilla.redhat.com/show_bug.cgi?id=1576481