redis-cluster redeploy cluster fail via kubernetes

Question

• i have use kubernetes statefulset object to create redis-cluster that cluster is fine first then i delete the statefulset object redeploy but type command "cluster nodes" to check cluster show cluster fail.

• redis-cluster.yaml

  • https://github.com/JayChanggithub/k8s-redis-cluster.git

• initial redis-cluster the cluster nodes is workable

  # the first time deploy redis-cluster
  $ kubectl apply -f redis-cluster.yaml
  
  # sts objects 
  $ kubectl get sts -n kube-ops
  redis-app      6/6     6m58s
  
  # create the cluster replicas as 1
  $ kubectl -n kube-ops exec -it redis-app-0 -- redis-cli --cluster create \
    -- cluster-replicas 1 $(kubectl get -n kube-ops pods -l app=redis -o \ 
    jsonpath='{range.items[*]}{.status.podIP}:6379 ' )
  # via one node of cluster
  $ kubectl exec -it redis-app-0 bash -n kube-ops
  
  > redis-cli -c
  > cluster nodes
  784f9d3930d7e00047fa6c8c53f3ea5c04bba4cb 10.244.3.171:6379@16379 slave 
  afc3aac801f1268ce8f7549deb3eb5eca92e178b 0 1603341990044 4 connected
  5ece7b219e0dbd5a17abfeb3af2a8ad847049891 10.244.0.85:6379@16379 master - 0 
  1603341991053 3 connected 10923-16383
  039aec28c71d0becc77d5703bb16775dd55796d9 10.244.0.86:6379@16379 slave 
  5ece7b219e0dbd5a17abfeb3af2a8ad847049891 0 1603341991000 6 connected
  2f30dc3d674e3dad98f42d2814e537a8bb458ed2 10.244.1.99:6379@16379 master - 0 
  1603341990000 2 connected 5461-10922
  103ece1c43de20ab7f995f2e23662551cb1b1606 10.244.1.100:6379@16379 slave 
  2f30dc3d674e3dad98f42d2814e537a8bb458ed2 0 1603341991053 5 connected
  afc3aac801f1268ce8f7549deb3eb5eca92e178b 10.244.3.170:6379@16379 myself,master 
  - 0 1603341988000 1 connected 0-5460
  

• redeploy the redis-cluster

  # delete before sts
  $ kubectl delete -f redis-cluster.yaml
  
  # redeploy redis-cluster
  $ kubectl apply -f redis-cluster.yaml
  
  # check the cluster status via one of cluster
  $ kubectl exec redis-app-0 -it bash -n kube-ops
  > redis-cli -c
  
  # output error as following
  > cluster nodes 
  

• output error

  root@redis-app-0:/data# redis-cli
  127.0.0.1:6379> cluster nodes
  c8cab83dedd37c9b84707758a2621d72dde4ca6b 10.244.1.78:6379@16379 master,fail? - 1603290611001 
  1603290609051 3 connected 10923-16383
  f4a3fa3d7b10903f32f6d46054f593541a82ee3c 10.244.1.79:6379@16379 slave,fail? 
  cfa04f400535980d21ec4a7db5320fb477c3c90d 1603290611613 1603290609051 2 connected
  7796faa8698d38d877657e488faa63f76306b531 10.244.0.71:6379@16379 slave,fail? 
  c8cab83dedd37c9b84707758a2621d72dde4ca6b 1603290609998 1603290609051 3 connected
  f4a56ba199724dd971d2c5d32079954c4f43a9e4 10.244.3.140:6379@16379 slave,fail? 
  22072f4f2d1c89f9f01d38a8c4b77a7d726c795e 1603290611613 1603290609051 1 connected
  cfa04f400535980d21ec4a7db5320fb477c3c90d 10.244.0.70:6379@16379 master,fail? - 1603290611613 
  1603290609051 2 connected 5461-10922
  22072f4f2d1c89f9f01d38a8c4b77a7d726c795e 10.244.3.142:6379@16379 myself,master - 0 1603290609051 1 
  connected 0-5460