Hi
tl;dr
I have a
redis cluster and I frequently take rdb dumps. Now I want to build a replica redis cluster on different set of nodes(different IP addresses) using these rdb files. Assume no TCP connectivity between old cluster and new cluster.
Please suggest me how can I achieve it.
Detailed problem:
I want to run redis cluster on kubernetes. I have tried redis cluster with persistent storage using Kuberneter statefulset and Kubernetes deployment but on cluster reboot a new set of IP addresses are assigned to my pods and they are not able to form cluster again. I have tried using kubernetes services with static ip and using hostname but redis internally stores pod IP and on reboot they are not able to form custer.
So, I am thinking about deleting node.conf file on reboot, create fresh cluster and restore rdb files. But I couldn't find any good reference for it. I found some good documents for redis cluster restore in Elastic Cache but they are specific to aws.
Has someone tried redis cluster on kuernetes with cluster reboot resiliency?
Is there any better way of restoring cluster without forming a new cluster in case of IP address change?
Link to my redis cluster using Kubernetes deployment configs:
https://github.com/zuxqoj/kubernetes-redis-cluster/blob/master/README-using-deployment.mdLink to my redis cluster using Kubernetes statefulset configs:
https://github.com/zuxqoj/kubernetes-redis-cluster/blob/master/README-using-statefulset.mdThanks.