Subsequent tflocal apply runs break ALB/ECS cluster

Platform
1

I’m deploying an ECS cluster behind an ALB to LocalStack. This deploy works correctly to the live AWS cloud, and it works correctly when deploying to a fresh LocalStack instance. However on subsequent runs of tflocal apply, despite no changes to template files there is a proposed change to my ALB listener to change my port from 4566 to 65000, despite 65000 being the original port the ALB is set to in the template files. It seems that on the first deploy, LocalStack is somehow changing the listener port from 65000 to 4566 and when subsequent deploys are applied the app breaks and the listener is unreachable.

Any help?

2

Hey @chris.mc, could you share the tf file?

3

i’d love to - but it seems .tf files aren’t an authorized filetype for an attachment. is there a recommended method of sharing on this forum that isn’t copy/pasting the contents of the files here?

4

any tips here? i can take a bunch of screenshots of the template files if pastebin is my best option here

5

more info in this issue bug: settings for ALB and ECS in template files are overwritten by LocalStack on deploy · Issue #10766 · localstack/localstack · GitHub