I've been running bots on Railway for 2 years and I've never had any issues with latency

Do you have a database on Railway that your bot is using?

yep, Postgres

Can you make sure they're on the same region and that you're connecting to postgres through the private network?

that would just be DATABASE_URL?

exactly, to be sure you can check if it uses a railway.internal domain

right, i've been using that one

awesome, and the region?

project and db are US West

on the exact same area in US-west?

yea

they're both US West Oregon ?

<:sus_apple:1032821016319234068>

yep

Alright, then your service's connection to the database is as fast as can be (low single digit ms range)
beyond optimizing that connection, there isn't anything else you can do to speed up things on Railway's side, the rest is your app

My suggestion would be to place more metrics in-app so you can figure out what exactly is slowing your app down, and you can work from there

this is one of my bot's running on Railway

i had to migrate from SQLite to Postgres, besides that and the hardware I would expect them to behave the exact same

can try some logging though ^

im kinda confused from reading, are you running part of the bot on your pc, or you are saying there is a 10 second delay from discord command to bot processing it on railway?

10 second delay is only on railway, if i host it locally latency is minimal

ohhhh

what library are you using?

discordpy

hmm, I was wondering if it was time server stuff, but prob not an issue in discordpy

do you get that delay on all commands? even a simple ping command?

good question

seems its just the memory-intensive commands

hmm

though it doesnt respond to some other ones <:clueless:1167482383084290078>

huh

just how great is your pc lmao

not too great lol

yeah…. ok

only 32gb vs 8gb but not making the difference in this case

agreed

what do these memory intensive commands do?

make a bunch of web requests, stitch tiles together and render an image, do a projection and send it to the channel

im interested also in that some commands do not get a response

probably a bug on my part

oh.

so it did not work fully on your pc

i didnt do the postgres branch on my local

ah

ill compare on my local now

seems like a consistent 2-4s

so at least partially my fault lol

hmmm

is the extra time just network request times?

that would be my guess

might just try some sqlite stuff

hmm