Build logs would just stop halfway through with no error message:
Pushing [=============================> ] 2.756GB/4.605GB bf5f31aa5c3f

Pushing [=============================> ] 2.759GB/4.605GB bf5f31aa5c3f

Pushing [==============================> ] 2.763GB/4.605GB bf5f31aa5c3f

Pushing [==============================> ] 2.768GB/4.605GB bf5f31aa5c3f

Pushing [==============================> ] 2.772GB/4.605GB bf5f31aa5c3f

Ignore my issue, it seems to have solved itself. No code changes, just a commit to restart deployment.

I think this has happened before a few times, builds are exiting unexpectedly. If you want to debug, here is the path URL.
cdaa4865-8678-4abe-acde-066cf95140bc/service/339d4695-6dcf-4037-9d8f-588e1b9095ac?id=dc145ab7-f16c-49f0-b921-6c5d16f08775

Known issue, but it does not seem to effect the deployment from going out, it just reduces observability, Railway has a fix in the pipeline for those frozen logs.

Additionally, may I ask what kind of app you're deploying that is resulting in a 4.6GB image?

This is still happening occasionally. Just had another two builds fail randomly.

It's a pretty huge next.js app with some images sourced locally (I know, not a great idea but we're still in active Dev)

I've figured out what the issue is. I wasn't aware that bun was now the default node runtime (when there's a bun lockfile?) and when it attempts to build and fails, it doesn't fail the build step at all and just goes straight to deploy which fails.

This doesn't seem to happen with npm run build.

This DOES affect the current deployment because it takes it offline and replaces it with the new deployment meaning I just get a "Application failed to respond" when it should just fail the build.