Since approximately 2025-11-25 05:30 UTC, all new deployments are failing with the error "Failed to build an image". No detailed build logs are provided - only the following appears:

[inf] scheduling build on Metal builder "builder-zdnszj"
[inf] fetched snapshot sha256:... (12 MB bytes)
[inf] [snapshot] fetching snapshot, complete 12 MB

Then the build immediately fails without any error details.

What I've tried:

1. Rollback to last successful deployment - Works fine (commit 2594746)

2. Deploy using railway up from apps/server directory - Fails

3. Deploy identical code that was previously successful - Fails

4. Removed new files (.sql migration file) - Still fails

5. Reverted all code changes to last working commit - Still fails when using railway up

Environment:

• Railway CLI version: 4.11.1

• Builder: Nixpacks (nixpacks.toml)

• Node version: 18

• Root directory: apps/server (monorepo structure)

Recent failed deployment IDs:

• cc9d567f-91ea-48b4-a231-24ae9b4dee41

• 019d60cb-35e4-4208-afcf-36a3208cabbd

• 57056ecc-d29c-4d63-a1d7-3cb1f665eed2

• d992bbbe-8e00-4e87-8bbc-a1010b5bcd31

Last successful deployment:

• Commit: 259474633104cb314b1c75928d3774088775645b

• Time: ~2025-11-25 05:29 UTC

Questions:

1. Why are there no detailed build logs showing the actual error?

2. Is there an issue with the Metal builder "builder-zdnszj"?

3. Why does rollback work but new deployments fail, even with identical code?

Please investigate and provide detailed error logs.

Thank you.