Eugene Low Migrating Standalone MongoDb Database to Kubernetes on AWS
While configuring the connection between the applications in K8s Pods, I found out that there would be lots of custom work if I want to use an external database for the pods. Therefore I seek for a better solution.
I know that there is few Helm Charts (by Bitnami) that will help me setup the MongoDb database on frontend cluster, but I am looking for a solution to migrate the data from one server to the cluster.
Morever the Helm Chart have 31 templates! That is too much for any workload, therefore I wish to check the rendered result before moving towards deploying the charts onto the cluster.
The Bitnami template contains a sets of customizable values one of which caught my eye:
persistence.existingClaim: Provide an existing PersistentVolumeClaim (only when architecture=standalone)
I know that I will need to commit my current database data into the PersistentVolumeClaim and setup the database with credential. So when the MongoDb pods are running, existing data can be in the database.
Multinode solution
The database pod can be running in any of the nodes, the PresistentVolume feature(or I should say “requirement”) that allow node have access to the presistent volume can be helpful.
My solution is to use the AWS EBS Multi-Attach to attempt to attach the current database volume to all the nodes on the cluster.