How to run PG commands on a Railway database

Just like the last post, I love Railway, and I always use it for my databases.

I was doing some sus work on my production DB for a client’s website, so ofc I didn’t wanna nuke it, so I thought well let’s back it up and then I’ll restore if shit goes south.

Now, ofc I shouldn’t be doing this, I should have a separate experimental/development PG instance where I do this, but this is a site that isn’t even live + I’m the only one working there, plus testing directly in prod gives you the maximum speed required to escape humanity’s biological limitations.

That being said, here’s how you run pg_dump on a PG instance through the Railway proxy:

pg_dump "insert public connection string here" -f file-name.sql

Copy and paste the DATABASE_PUBLIC_URL, keep the quotation marks. That easy, make sure you got postgres installed on your system, super easy to do using homebrew:

brew install postgres

If you get an error about version mismatch from pg_dump, check what version you have installed:

pg_dump --version

If your version is older than the one Railway is using (the error will literally say this), install the matching version, for me it was version 16.10, so I used:

brew install postgres@16

Follow the post-install instructions homebrew provides and you should be good to go.

For psql, it’s also very simple:

psql "insert your connection string here" < your-sql-file.sql

That’s it, that’s all there is to it, have fun.

Btw, I learnt this while creating that development instance of PG and copying the production instance’s data to the dev instance, gonna go live, can’t blow prod up.

Adios, amigos 👋