Hey there! We've found the following might help you get unblocked faster:

• 🧵 URGENT: Resizing volume of Typesense instance stuck forever

• 🧵 Database unavailable – stuck on "Resizing volume…"

• 🧵 Resizing volume failed

If you find the answer from one of these, please let us know by solving the thread!

The resizing seems to be done, but now the postresql server keeps crashing with:

2026-01-10 18:51:55.902 UTC [5] FATAL: bogus data in lock file "postmaster.pid": ""

Certificate will not expire

PostgreSQL Database directory appears to contain a database; Skipping initialization

Mounting volume on: /var/lib/containers/railwayapp/bind-mounts/XXXX

I fixed it by setting this as start command: rm -f /var/lib/postgresql/data/pgdata/postmaster.pid and then reverting to my normal start command.

To be completely honest: these quirks makes me consider switching away from Railway. All these pains really shouldn't be there just to grow a volume, and it's not the first time I've had to dive deep into this community and Google to fix something. I pay a relatively big premium for Railway over something like Coolify to not having to fix these things myself, as a complete rookie, with half a day's downtime.

Hi,

So glad you were able to solve your issue. We really appreciate you being a Railway customer. Unfortunately this issue relates to Postgres and can occur regardless of what platform you are on. We aim to make Railway easy to use, but we cannot solve all third-party software complexities unfortunately.

Let us know if we can be of further help.

Thanks, but how can a volume resizing that didn't work on multiple tries and took 2-3 hours before getting a "result" and seemingly corrupted something in the postgres not be platform related? I don't think it's postgres that handles the disk volume?

Again, sorry for your frustration -- this happens if/when Postgres doesn't shut down cleanly. In this case, it was terminated before it could do a graceful shutdown.

PostgreSQL uses postmaster.pid as a lock file to prevent multiple instances from running. When PostgreSQL starts, it writes its process ID to this file. On clean shutdown, it removes the file. Since it was killed during the resize, the file remained.

To help prevent this in the future, you can add a railway.json with a longer shutdown grace period:

json

{
  "deploy": {
    "stopGracePeriodSeconds": 30
  }
}

Let us know if you need more help.