(assets): Uploaded file must be a non-empty zip

[patrick-tresp]

We are facing the issue, that we can not deploy any environment ( production, staging, development, test ) with cdk deploy anymore.

The error that raised:

Resource handler returned message: "Uploaded file must be a non-empty zip (Service: Lambda, Status Code: 400, Request ID: 11cbd3f4-e5a4-4b7d-a10a-15da3
58f4b64, Extended Request ID: null)" (RequestToken: a6d89298-48b5-44ce-215a-14bd022bca02, HandlerErrorCode: InvalidRequest)

We can see that cdk.out folder is filled with aseets.[assets-id] folders, but no zip folders are created.

Also interesting is the fact, that if we create a fresh account on aws, bootstrap it and deploy there, everything works as expected for exact one time. Any deployment afterwards ( of course with little changes to force an update ) fails with error above.

This happens on local macOS machine as well as previous working github actions with ubuntu/latest OS.

Reproduction Steps

run cdk deploy STACKNAME --profile [aws-profile]

Note: [aws-profile] is set in ' ~/.aws/credentials`

What did you expect to happen?

Assets are beeing created in cdk.out folder and zipped for an upload to S3 Bucket to execute Cloudformation deployment.

What actually happened?

No zip files are created in cdk.out folder.

Environment

• CDK CLI Version : aws-cli/2.2.21 Python/3.8.8 Darwin/20.5.0 exe/x86_64 prompt/off
• Framework Version:
• Node.js Version: v14.17.3
• OS : MacOs 11.4
• Language (Version): TypeScript (3.9.7)

Other

---

This is 🐛 Bug Report

[mattcui]

I met the similar problem with my golang based lambda function, but I can see the built main file in the asset folder, like cdk.out/asset.08e3e0605a610fe153bcd368387d926e0e5ef54458413fb7232b708045dcaf63/main (note that it's not a .zip file), and I can see the zipped main file in s3 stage bucket.

[patrick-tresp]

@mattcui did you find any solution on that?
it seems like it brings our whole infrastructure to a sudden stop.

[mattcui]

@patrick-tresp not yet, I have been struggling for 2 days, no any progress so far.

[mattcui]

@patrick-tresp Today, I retried the deployment, it's successful this time, looked that something was fixed perhaps from server-side. You could have a try.

[patrick-tresp]

Hey @mattcui , thanks for the notice.
Unfortunately it still runs into same issue BUT a fresh provisioned account does NOT run into this error.

Will run several deployments now to verify and move resources to that account. :(

[patrick-tresp]

Well after moving all stacks and ressource to the other account, deployment fails also here after a couple passing deployments…

What is going on here…

[patrick-tresp]

Hey @mattcui ,

try this out: empty your s3 bucket from toolkit cdktoolkit-stagingbucket-xxxxx.
After emptying this, we could deploy again without running into this error.

We also added this to your github action file with

aws s3 rm s3://cdktoolkit-stagingbucket-xxxxx --recursive

Hope that helps you as well.

Cheers,
Patrick

[mattcui]

@patrick-tresp Thanks, we haven't had this problem for days, I am not sure what caused the problem, and why it disappeared.

[patrick-tresp]

AWS Magic …

[github-actions]

⚠️COMMENT VISIBILITY WARNING⚠️

Comments on closed issues are hard for our team to see.
If you need more assistance, please either tag a team member or open a new issue that references this one.
If you wish to keep having a conversation with other community members under this issue feel free to do so.