frequent job failures due to artifact upload failures

The cloud team has recently (since approximately 2025-03-23) started experiencing pipeline failures due to timeouts when uploading job artifacts to salsa's internal storage. The failure are accompanied by the following in the job logs:

Uploading artifacts for successful job
15:17
Uploading artifacts...
*.build.json: found 1 matching artifact files and directories 
*.info: found 1 matching artifact files and directories 
*.tar.xz: found 1 matching artifact files and directories 
WARNING: Uploading artifacts as "archive" to coordinator... 504 Gateway Timeout  id=7327341 responseStatus=504 Gateway Timeout status=504 token=64_1zJe1G
WARNING: Retrying...                                context=artifacts-uploader error=invalid argument
WARNING: Uploading artifacts as "archive" to coordinator... 504 Gateway Timeout  id=7327341 responseStatus=504 Gateway Timeout status=504 token=64_1zJe1G
WARNING: Retrying...                                context=artifacts-uploader error=invalid argument
WARNING: Uploading artifacts as "archive" to coordinator... 502 Bad Gateway  id=7327341 responseStatus=502 Bad Gateway status=502 token=64_1zJe1G
FATAL: invalid argument

You can see a complete example at https://salsa.debian.org/cloud-admin-team/debian-cloud-images-daily/-/jobs/7327341 or find others by looking for recent build stage failures in that pipeline.