Postgres slow/timeout
fra
HOBBYOP

4 months ago

Hello,

I'm still in the free trial to test railway infrastructure, atm my app is still in dev, nobody use it but me, the metric are flat almost to 0 but when I run my e2e tests that writes ~150 rows in a few minutes I can see that sometime the app timeout, and the reson seems to be postgres, the only error log I can find in my stack is this pg log:

2025-08-14 17:30:01.315 UTC [74] LOG:  checkpoint complete: wrote 182 buffers (1.1%); 0 WAL file(s) added, 0 removed, 0 recycled; write=17.855 s, sync=0.022 s, total=19.032 s; sync files=134, longest=0.008 s, average=0.001 s; distance=768 kB, estimate=768 kB; lsn=0/2FAB2E0, redo lsn=0/2F8FC60

Also, I have one more question, am I right when I say that these pg instances are not managed? Do you offer a managed postgres solution?

thanks

Solved

7 Replies

fra
HOBBYOP

4 months ago

0efc8ea7-b963-4d84-b9de-c6e45c058c0f


fra
HOBBYOP

4 months ago

btw, this might mean I'm on metal, I'm happy to stay on metal and somehow help if I can


fra
HOBBYOP

4 months ago

any update on this?


medim
MODERATOR

4 months ago

Hi! The instances are not managed instances, it just deploys the postgres image and attaches a volume to it.


medim
MODERATOR

4 months ago

I will escalate this to the team, I wonder if it's related to the I/O issues some instances are having


medim
MODERATOR

4 months ago

!t


medim
MODERATOR

4 months ago

This thread has been escalated to the Railway team.

Status changed to Awaiting Railway Response medim 4 months ago


fra
HOBBYOP

4 months ago

thanks, is not super urgent, but I'm looking to go live soon with my app, I really like railway but if the db has this issues with just me trying the app I'm not sure if it is a good idea to go live


christian
EMPLOYEE

3 months ago

Confirming that database instances on Railway are not fully managed, but offers features like backups: https://docs.railway.com/reference/backups


Status changed to Awaiting User Response Railway 4 months ago


christian
EMPLOYEE

3 months ago

Looking at your application HTTP logs, you have some 499 http status codes, so these appear to be cases where the client is closing the connection, and not an issue between your application and database, or the performance of the database.


fra
HOBBYOP

3 months ago

yep but I think the client close the connection because the db doesn't respond in time, the only errors I can see are in the db, where in the logs I can see WAL checkpoints taking even 17s to write, this seems similar to other people issues with postgres


Status changed to Awaiting Railway Response Railway 4 months ago


brody
EMPLOYEE

3 months ago

Hello,

17-second intervals for WAL writes are normal in Postgres. Postgres doesn't write WAL to disk immediately - it spreads writes out over determined time periods for efficiency, so that metric alone isn't indicative of a performance problem.

To identify what's causing your application slowdown, I'd recommend adding tracing to pinpoint where the performance bottleneck is actually occurring.

Best,

Brody


Status changed to Awaiting User Response Railway 3 months ago


fra
HOBBYOP

3 months ago

yeah I've done some research after writing and found out my ignorance on this bit, I'll do some more tests to understand if it's my app or not


Status changed to Awaiting Railway Response Railway 3 months ago


brody
EMPLOYEE

3 months ago

Sounds good, let us know what you find!


Status changed to Awaiting User Response Railway 3 months ago


Railway
BOT

3 months ago

This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!

Status changed to Solved Railway 3 months ago


Loading...