a month ago
Deployment ID: ffc02126-9db1-42dd-8377-7bc64d25c631 (the latest failed build)
Service: newguard (frontend)
Problem: Build consistently fails. Diagnosis says Railway is using Railpack instead of Dockerfile despite:
builder=DOCKERFILE in railway.json
Dashboard set to Dockerfile builder
RAILWAY_DOCKERFILE_PATH variable set
All build args (ARG_VITE_*) now configured
Blocker: Workspace log query limit (48 concurrent) is preventing us from seeing actual build logs
Request: Pull server-side build logs for that deployment ID and diagnose why the Docker build is failing
Pinned Solution
a month ago
Hey! Make sure railway.json is at root and not in a nested route, also make sure RAILWAY_DOCKERFILE_PATH matches actual file location.
Since you can't see build logs, can you build locally with the same args to see if it works?
2 Replies
a month ago
This thread has been opened as a public bounty so the community can help solve it. The thread and any further activity are now visible to everyone.
Status changed to Open Railway • about 2 months ago
a month ago
Hey! Make sure railway.json is at root and not in a nested route, also make sure RAILWAY_DOCKERFILE_PATH matches actual file location.
Since you can't see build logs, can you build locally with the same args to see if it works?
a month ago
Thanks Mr. Mod,
Support's suggestion — "build locally with the same args" — is what cracked it. I ran docker build with the exact ARG_ values: it failed on qa, and succeeded on production. The stuck "48 concurrent log queries" limiter had hidden the real error the whole time, which is why it looked like infra. Lesson logged: when Railway's logs are walled off, reproduce the build locally with the same build args first. (There were genuinely two stacked issues — the Dockerfile builder not activating early on, then this env value — and the log jam masked both.)
- Verified the ARG_* vars and set the Docker builder path via CLI.
- Forced a real build → it failed (~30s), same as support's attempts — ruling out their build-arg theory.
- Reproduced locally → caught VITE_ENVIRONMENT=qa as invalid; confirmed production builds clean.
- Set VITE_ENVIRONMENT/ARG_VITE_ENVIRONMENT=production via CLI, forced a rebuild → SUCCESS, site live.
Status changed to Solved medim • about 2 months ago