Arbitrarily failing CI/CD: "Deployment is already in progress."

[narrow musk]

Has anyone else been experiencing their CI/CD deploys fail with error message containing "Deployment is already in progress" even though there is no other deployment in progress?

[narrow musk]

(With cli 10.8.1)

[torn ivy]

Yep! I have and it is driving me nuts

CLI version 11.0.5

The root stack in the Cloudformation console is "UPDATE_COMPLETE", yet Amplify CLI insists that deployment is in progress 🤦‍♂️ did you have any luck with a fix?

[ocean folio]

If I remember correctly, I have faced similar issue and then it was caused by aborting amplify push midway.

Do you see a deployment-state.json file in your deployment bucket? If so, removing it should unblock you.
https://github.com/aws-amplify/amplify-cli/issues/9665

[narrow musk]

Thanks for your replies @torn ivy , @ocean folio . The problem went away but I'm not sure what fixed it in the end. Since then I had upgraded cli to 11.0.X and now I'm on 11.0.5.

If it comes back I'll try the solution you linked to, @ocean folio .

[torn ivy]

Thanks @ocean folio! I deleted my stack and started over, but I might not have that luxury in the future so I will remember your solution. Cheers guys!

[autumn hull]

Do you have your source repository connected to Amplify with CI/CD enabled? If you push code to your repository it will trigger a build, if that build is running when you try to amplify push you will receive an error about a build already in progress.

[narrow musk]

I do, but no build was running cloud-side when the issue arose.

[severe creek]

I've just come across this. It looks like there is a difference between the build timeout for amplify (set in build settings on the console, defaulted to 30 minutes) and event timeout for cloudformation (seems to be 2 hours).

In my case the amplify build is failing after 30 minutes, so I thought I could push again, but cloudformation is still attempting a delete. So, I get this error.