https://www.traditionrolex.com/36

vRealize Automation 6.2 moving to Internal replicated vPostgres

When delivering a vRealize Automation 6.1 and prior in a distributed install, the recommendation was always to have an external vPostgres instance. This has changed in vRealize Automation 6.2 and it’s briefly touched upon in the vRA 6.2 reference architecture found at: vRealize Automation 6.2 reference Architecture. I wanted to make sure to highlight this, as many have missed this change and others are still wondering how it can be implemented or why this change. It’s important to mention as well, you can use this replication architecture for any version of vCAC/vRA that’s higher than 6.0.1.

Let’s start by highlighting the change, we are moving from having a separate/external vPostgres  machine to host the vPostgres (being a vRA appliance that is re-purposed to be used as vPostgres or a full vPostgres install) to using the internal built-in vRA Appliance vPostgres and replicate across the two vRA Appliances. The below image demonstrate this change:

vRA internal vPostgres replication
So what do you get of moving to this new model? Actually there is few advantages of implementing your vRA 6.2 this way, below is couple of them.

1- You have less machines to manage by eliminating the extra one or two external vPostgres machines you needed in the 6.1 reference architecture setup.

2- You have a replicated DB without the overhead of extra two machines.

3- You will be more future proof for upcoming release, as you will be following the current reference architecture.

How do I implement vRA internal vPostgres replication?

Below I will describe the steps at a very high level, as there is a KB article that do a great job in describing this in a step by step fashion that I will point out to at the end of this post:

1- Deploy two VMware vRealize Automation appliances(Don’t configure them yet), and make sure they are resolvable through DNS.
2- Create a DNS entry for the vPostgres VIP
3- Configure the database VIP using a load balancer or DNS
4- Add an extra 20GB disk to each vRA appliance
4- Enable the replication on both appliances using the scripts included in the 2108923_dbCluster.tar package
5- Perform a test failover (Appliance A to Appliance B)
6- Perform a test failback (Appliance B to Appliance A)

You can find the step by step instruction in KB2108923: Configuring appliance database server for replication in VMware vRealize Automation Center virtual appliance (2108923)

Comments

  1. toastymel says

    Thanks for posting this! I was curious if you knew where to find the 2108923_dbCluster.tar file? The KB Article nor the guides show where to actually obtain this file.

  2. Toastymel, Actually they have the download link right at the bottom for that KB.

Trackbacks

  1. […] Internal Replicated vPostgres This is a most excellent change, and I agree with the author of this article that not everyone fully realizes this yet, and I think they should.  At 6.0.1 and later you can […]

Speak Your Mind

*