500 Internal Server Error

Once a repository configured and deployed to workers with custom subdomain, and username/password credentials.

Pushing a Docker image with `gradle jib` for a standard Spring boot app is sometimes failing.

Most the time it succeeds and something I get the following error at random:

```
com.google.cloud.tools.jib.plugins.common.BuildStepsExecutionException: 500 Internal Server Error
PUT https://cr.project.acme.com/v2/my-image-staging/manifests/sha256:2c3c85c056561b287636deb0fc17818ad5624bae1b3e018550f9224dfe4540fe
{"errors":[{"code":"INTERNAL_ERROR","message":"internal error","detail":null}]}
```

Looking at Workers logs, here are errors I get:

https://gist.github.com/lpellegr/3e75362d6bb8b31c733a2b00267a5a44

Cloudflare status does not show any outage.

The issue seems to appear at "random". However, I noticed that if I deployed a new version of the serverless registry to worker following a wrangler.toml configuration update, then the next few uploads seems to often fail.

Any idea what's the cause?

Start by reviewing the Workers errors in the linked gist and the deployment settings in wrangler.toml, then reproduce Docker image pushes with Gradle Jib after deploying a new serverless registry version. Done means identifying the cause of the intermittent manifest-upload 500 errors and verifying that repeated uploads succeed after deployment.

Written by the indexing model from the issue text.