Hello, I am trying to create two CockroachDB clusters in two different Amazon clusters and using helm to deploy it. Under the covers I am using Skupper to create a tunneling between both clusters at service level.
So this is whatI do. I create a CockroachDB cluster with defaults on Cluster1. Everything is deployed correctly.
Then from Cluster2 I run the
kubectl run cockroachdb -it \ --image=cockroachdb/cockroach:v20.1.1 \ --rm \ --restart=Never \ -- sql \ --insecure \ --host=skupper-cockroachdb-public
And it works and I can create DBs and tables and so on, so the communication between clusters is possible and using tunneling I am able to create a DB from cluster 2 to cluster 1 deployed cockroachDB.
But then I want to make things to another level and I want to extend the CockroackDB cluster to cluster 2.
So I use Helm again and in the join section, I add the skupper-cockroachdb-public host (which is a kubernetes service that points to the three pods deployed on cluster 1) as it is the hostname that is used in tunneling to refer to the cockroach pods. But then the three nodes created in cluster 2 are not ready, they are deployed but readiness probe retruns a 503. I think I am pretty close but I do not know exactly what is the latest step before making it run.
I am using an insecure installation.
Any help on this?
Thank you very much