From the deploy log, can you see which resource got this error?

We added on our app

const removalPolicy = app.stage === 'prod'
    ? RemovalPolicy.RETAIN
    : RemovalPolicy.DESTROY;

  app.setDefaultRemovalPolicy(removalPolicy);

When seed deploys our dev environment, now it crashes everytime. It deployed perfectly friday before we added the default removal policy

I added:

s3Bucket: {
        autoDeleteObjects: scope.stage !== 'prod',
        removalPolicy: scope.stage === 'prod'
          ? RemovalPolicy.RETAIN
          : RemovalPolicy.DESTROY,
      },

to our

ReactStaticSite

but it does not seem to do the job.

I thing I understood why it is in a deadlock: • My react site was named ‘foo’ • We deployed and it created a bucket …-foo with a custom cname and an entry in route 53. • We (Stupidly) changed the name to ‘bar’ • It didn’t delete the bucket …-foo, when it tries to create …-bar it fails because the CNAME is already in use • Then it tries to delete the freshly created bucket. And it fails because there is objects in it. It shouldn’t fail because of objects as we set the autoDeleteObjects property

So by setting the

autoDeleteObjects

, a custom resource is created backed by a Lambda function that removes all the files in the bucket. My guess for the scenario you mentioned above is that, after the

xxxx-bar

bucket was created, the CNAME record failed BEFORE the autoDeleteObject custom resource was created.

This unfortunately is a flaw with how custom resource works.