'Pods are not rescheduling to failure node ,when node come alive

My situation is ,I have 5 nodes having K8 cluster .Initially pods are distributed across the 5 node. Sometime we need to restart particular node server. Then that node goes down and and pod will created on another node. But once failed/down node comes up ,no pods are creating in it automatically, as replica number already reached .We need all node have minimum 1 pods to run .Could please help on this

kubernetes


Solution 1:[1]

We need all node have minimum 1 pods to run

Instead of running the Deployment, you can run the Daemon set if you want to run a minimum of one 1 pod to run on the node.

So if anytime new comes to the cluster it will have one replica running of your POD.

Read more daemon set : https://kubernetes.io/docs/concepts/workloads/controllers/daemonset/

You can spread also replicas if running as deployment however it would be tricky to manage with Affinity, POD topology spread

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source
Solution 1

Related Questions