Production Redis cluster with sharding in Kubernetes - kubernetes

Question

I have tried Redis stable 'helm' chart to deploy a Redis cluster with 1 master and 3 slaves that replicates data written to master. But it is a single point of failure - I deleted the master, no new pod of master was recreated. Also, the chart does not support data partitions (sharding).
EDIT: I have created a Redis cluster using helm redis-ha chart, but there is no option to have sharding.
Aren't there Redis helm charts to deploy a production ready HA cluster that supports partitions (sharding)? Can you point me to resources I can use to setup a manageable Redis cluster? Primarily, my Redis is used for data caching, message processing & streaming.

Answer 1

You need Redis Sentinel.
If helm is an option, these links can be of help:
https://github.com/helm/charts/tree/master/stable/redis (notice the sentinel-related configuration parameters)
https://github.com/helm/charts/tree/master/stable/redis-ha

Answer 2

Here's a fine tutorial about setting up a 3 master / 3 slaves redis cluster with partitioning. It's targetted at Rancher, but that's optional, I just tested it on Azure Kubernetes Service and it works fine.
First, apply this yaml (ConfigMap, StatefulSet with 6 replicas, Service):
apiVersion: v1
kind: ConfigMap
metadata:
name: redis-cluster
data:
update-node.sh: |
#!/bin/sh
REDIS_NODES="/data/nodes.conf"
sed -i -e "/myself/ s/[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}/${POD_IP}/" ${REDIS_NODES}
exec "$#"
redis.conf: |+
cluster-enabled yes
cluster-require-full-coverage no
cluster-node-timeout 15000
cluster-config-file /data/nodes.conf
cluster-migration-barrier 1
appendonly yes
protected-mode no
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: redis-cluster
spec:
serviceName: redis-cluster
replicas: 6
selector:
matchLabels:
app: redis-cluster
template:
metadata:
labels:
app: redis-cluster
spec:
containers:
- name: redis
image: redis:5.0.1-alpine
ports:
- containerPort: 6379
name: client
- containerPort: 16379
name: gossip
command: ["/conf/update-node.sh", "redis-server", "/conf/redis.conf"]
env:
- name: POD_IP
valueFrom:
fieldRef:
fieldPath: status.podIP
volumeMounts:
- name: conf
mountPath: /conf
readOnly: false
- name: data
mountPath: /data
readOnly: false
volumes:
- name: conf
configMap:
name: redis-cluster
defaultMode: 0755
volumeClaimTemplates:
- metadata:
name: data
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 1Gi
---
apiVersion: v1
kind: Service
metadata:
name: redis-cluster
spec:
type: ClusterIP
ports:
- port: 6379
targetPort: 6379
name: client
- port: 16379
targetPort: 16379
name: gossip
selector:
app: redis-cluster
And the next step is to run this script to form the cluster (you will have to enter "yes" once interactively):
kubectl exec -it redis-cluster-0 -- redis-cli --cluster create --cluster-replicas 1 $(kubectl get pods -l app=redis-cluster -o jsonpath='{range.items[*]}{.status.podIP}:6379 ')
for x in $(seq 0 5); do echo "redis-cluster-$x"; kubectl exec redis-cluster-$x -- redis-cli role; echo; done