Some workflow jobs consistently fail during Release Build

[pulumi-bot]

Workflow Failure

Release Build has failed. See the list of failures below:

• 2024-10-04T07:02:10.000Z

[justinvp]

Succeeded on re-run, but it seems like every time there's a release some of the jobs fail and need to be re-run.

Can we avoid it?

[julienp]

Noting some recent failures to help with figuring out the best approach to make this more reliable.

• https://github.com/pulumi/pulumi-docker-containers/actions/runs/11174698118/attempts/1 failed due to a failed npm download for kitchen sink during a test (TestPulumiTemplateTests/aws-typescript), and the Debian images failed during manifest creation due to a network timeout when reading from the docker registry.

• https://github.com/pulumi/pulumi-docker-containers/actions/runs/11069527319/attempts/1 failed due to a network error while installing a specific node version during a test. (TestKitchenSinkLanguageVersions/node-22.5.1)