Resiliency between Cluster in One region


#1

Hi Team,

We have our application setup in one of AWS region. In this region , we have two Couchbase cluster (detail below , each cluster have 3 nodes) in different AZ. Our problem is our application is not resilience to handle the Cluster failure because we have mentioned connection string in application configuration which point to one cluster at a time. If cluster got failed where our application is pointing , we have to changes the configuration in OpenShift and redeploy the pods.

Cluster C1 :- couchbase://10.109.0.xxx,10.109.0.xxx,10.109.0.xxx

Cluster C2 :- couchbase://10.109.0.xxx,10.109.0.xxx,10.109.0.xxx

We want automatic Cluster fail-over management
Question :- Do we need a code change for this If yes any reference of node JS.
How I came to know the Cluster is down. Did below line will throw error
this.cluster = new Couchbase.Cluster(couchbaseUri)
Or It’s some configuration is required on Cluster or Infrastructure level.


#2

A more typical way to configure would be to create a single cluster; but use Rack/Zone awareness to define two server groups (one per availability zone). Then you are protected from the loss of a complete AZ; and don’t need to handle switching the application between clusters.


#3

Thank Drigby , but Rack/Zone awareness is architecture changes because we already setup our infrastructure as I mentioned above.

Any suggest how can achieve the resiliency between cluster as per our infrastructure setup via code