Multiple regions?

[bmilsom]

Hi,

Just wondering if anyone has tried or had any success with deploying into multiple regions? And if so, what approach did you take?

While it would be simple enough to simply deploy the whole infrastructure to another region (and making sure that the runner labels don't conflict with labels in your existing deployment), I'm wondering if its possible to have the infrastructure deployed in region A able to talk to EC2 instances in region B, and whether this would also require the SQS queues and Lambdas to also be deployed into region B? What else might we need to think about for a multi-region deployment?

Just food for thought and brainstorming at this point. I might have a go at implementing this when/if we have time.
Thanks!

[bmilsom]

For future reference, the way we've implemented this was to have multiple environments (managed in GitHub Actions environments), and different .tfvars for each environment, where the region is specified.
The muli_runner module then deploys into the specified region, and we have a separate GitHub app for each environment.

There are other ways to do this, but this is what worked for us.