Best Practice: Standing up a copy of SMAX on managed EKS in the same VPC.

Greeting fellow SMAX Experts,

My company has SMAX 2022.11 running on a managed EKS environment. We have three different VPCs, one for DEV, STG, and Production (PRD). We work primarily out of DEV where we figure things out, then promote to Staging where we run through a series of tests, development documentation and training. Then roll into production. In a recent EKS upgrade to 1.24, we blew up Development which stopped 6-7 projects in its tracks. This was due to 3 add-ons that also needed to be upgraded as part of that EKS upgrade. Now we're about to upgrade to 2023.05, then from there work through the upgrades to 24.2 by August. Not wanting to blow up DEV and to avoid standing up another VPC for a POC, we're wanting to create a copy of DEV in the same VPC.

I'm thinking we can;

1. Create a new DB
2. Stand-up a new installation of SMAX on managed EKS (Different EKS cluster)
3. Restore the DB, and deploy packages to ensure forms, rules, etc are intact.

Another option is maybe;

1. Create a new DB, restore a backup from DEV.
2. Create a VolumeSnapshot of the SMAX environment and reconfigure to talk to the new database.

All of this is speculation, and I'm not sure of the best way to proceed. But we want to create this in our DEV VPC, without disrupting the running DEV instance. And we need this new instance (POC) to be an exact replica of DEV (In regards to data, customizations, forms, etc.) The objective is to iterate through the upgrades, to ensure we understand the steps, process, and don't break anything.

Any ideas out there on best practices, or things you've done that similar that worked?

Tags:

  • Verified Answer

    +1

    Finally got an official answer from internal expert on hosting SMAX in managed EKS. Their answers was that OpenText does not support running multiple copies of SMAX in seperate EKS clusters within the same VPC. This is due to references and connection details stored in the database. So the best practice and recommendation is to deploy a new VPC, setup a new EKS Cluster, and build out a new SMAX deployment. Then restore the database.