Stateful service time out if one of Node is down

I am  evaluating DataBase solution for my kubernetes cluster, So i installed percona-xtradb-cluster with three replica. To analyze failure situation i shutdown one node and meanwhile my test pods is reading data after every 5 seconds from DB Cluster but i observed out of 10 request 1 or 2 are getting timeout by saying “ERROR 2003 (HY000) Cant connect to MySQL Server on ‘pxc’ (110)” , if i further down another node which mean only one node is up error percentage is about 60% which out of 10 only 4 are successful

  • I tried by adding PodDisruptionBudget with minAvailable: 1
  • I tried by creating NodePort and infact this is also showing intermittent behavior mean if i directly try on specific node it some time time out.
Is it due to StatefulSet or something else is wrong?  If we have these issues how come its production ready?

One more Observation, if i shutdown primary node then it doesn’t work whole cluster is unavailable, infact  still two nodes are up

2020-06-17T13:18:49.027004Z 0 [Note] WSREP: (0c831c98, ‘tcp://0.0.0.0:4567’) reconnecting to d4b530ac (tcp://10.244.2.1:4567), attempt 24002020-06-17T13:18:52.027693Z 0 [Note] WSREP: (0c831c98, ‘tcp://0.0.0.0:4567’) connection to peer 00000000 with addr tcp://10.244.2.1:4567 timed out, no messages seen in PT3S2020-06-17T13:18:56.028533Z 0 [Note] WSREP: (0c831c98, ‘tcp://0.0.0.0:4567’) connection to peer 00000000 with addr tcp://10.244.2.1:4567 timed out, no messages seen in PT3S