Welcome! Please see the About page for a little more info on how this works.

0 votes
in Ions by
edited by

Hi! After renaming our lambdas in the ion-config.edn file [[1]], the ion deploy command fails with this cause:

   "Function not found:
    (Service: AWSLambda; Status Code: 404; Error Code: ResourceNotFoundException; Request ID: <...>; Proxy: null)",
   :class ResourceNotFoundException})}

I know that however we rename things in this config file, the deployment steps will delete and then re-create the lambdas under the new names.

If I rollback my config file, the deployment succeeds in recreating the old-named lambdas. But even then, it carries the instruction to delete the inexistent ...-yada-wss-... lambda, as if it thought it's really supposed to be there. What makes it think it's ever been successfully deployed?

If I use my new config file and I comment out the latter half, it succeeds in deploying the ...-yada-app-... lambda, and I see the step machine is asked to delete the inexistent ...-yada-wss-... lambda.

If I use my new config file without commenting any of its parts, we don't reach the step function because somehow, something expects the ...-yada-wss-... to be already there before calling the step function, which it isn't.

I've been trying for a few hours to find more logs about this. The error happens very quickly before the Step Function state machine could be hit, or the CreateCodeDeployDeployme lambda.

What's wrong about my wss lambda?

Thanks for any help


1 Answer

0 votes
edited by

Finally I tested a hunch that if I just used different lambda names, I'd be in the situation where the system will no longer expect this particular lambda to exist so bad that it fails deploying my app. This hunch turned out to be true.

My old name still appears in each deployment's list of lambdas to delete, but since I no more use the name, my deployments won't fail because of it.

I still wonder, though, how I could make it so that the system understands once for all that this lambda no more exists.