#Theres a really really annoying bug where the server is up running, but then it just stop working.

48 messages · Page 1 of 1 (latest)

tacit saddle
#

Basically as you can see on the time stamps on the the images, the server was up, then out of no where it just got shutdown. Every fucking time it runs for a couple hours then it just stop working and goes offline.

rocky anchorBOT
tacit saddle
#

Theres a really really annoying bug where the server is up running, but then it just stop working.

uneven maple
#

can you check the logs of the removed deployment, and a screenshot of your history would be helpful too

tacit saddle
uneven maple
#

That Starting Container and Stopping Container happens when the container is paused and unpaused respectively, the only way to do that through the UI is via serverless. Where the container is being Removed makes no sense though, you're saying you're not the one removing the deployments?

tacit saddle
#

Yep. I'm not the one removing the deployment, 100%

#

It literally just happened

#

Right now, i was watching my bot on my sec monitor, the bot stopped working, i open railway and the Repo was removed lol

#

This one doesn't even have the removed log

uneven maple
#

can you check the activity panel for clues?

tacit saddle
#

How?

#

Where do i see it

uneven maple
#

Top right corner, it'll look like a heartbeat icon

tacit saddle
#

Its not showing anything, it only shows when i manually turn the server on again

uneven maple
#

that is very strange, I have no idea what sorcery is causing this

tacit saddle
#

lmao

uneven maple
#

i'll escalate this to the team since I'm out of ideas

rocky anchorBOT
#

This thread has been escalated to the Railway team.

frank sage
#

We aren't removing the deployment on our end.

tacit saddle
#

Okay, so do you think it might be something on my code?

frank sage
#

Not sure, I don't have access to your code.

uneven maple
#

I originally thought your code might be exiting with code 0 (cause if it isn't code 0 it'd show that it crashed), but that only results in a Completed state, not a removed state, so to me idk if your code could cause this unless you are interacting with the Railway API to remove the deployment

uneven maple
frank sage
#

I was already watching.

uneven maple
tacit saddle
#

Nice emoji usage btw 🤣 . I will try to investage the code and see if i get some answers but i really i dont know

#

I had other project that i was using on railway

#

And it never happened

#

I mean, this bug of the repo being removed

#

o. o

#

Really weird

uneven maple
#

I tried some more things on my side, and yea I'm not able to reproduce this at all (without interacting with the API)
i'm fully out of ideas

tacit saddle
#

I took a look at the bot codebase.

There is no Railway API usage, no GitHub/GitLab API usage, no child_process/exec/spawn, and no code path that could remove or detach a deployment/repository from Railway.

The app can only:

  • start the HTTP server
  • exit normally in scan mode
  • fail startup and set process.exitCode = 1 on fatal errors

That could explain Completed/Crashed states, but not a service showing “There is no active deployment for this service” and asking to manually reconnect the repo.

The only external network calls in the backend are to Hyperliquid metadata and public icon URLs, both unauthenticated and unrelated to Railway.

Theres no evidence whatsover that the code is the reason for the repo showing that there is no active deployment. And btw, it just happened again sadge . I even went to github to see if the server was actually deploied in the last commit that i send and it was

tacit saddle
misty ruin
#

should i help @tacit saddle i guess main sir brody will handle

frank sage
#

Help away, there is nothing for me to handle here, the issue is not on our end.

misty ruin
#

to fix it

#

so the issue is railway won't allow any thing which is posted as a bug likelly a exploit

#

so might be that issue

but i can see your build might have same issue what i told

old dependencies

#

or issue with a debug build not a production one

i must say try to update via npm install fund etc etc

frank sage
#

Please disengage if you are going to make things up.

misty ruin
frank sage
#

Yes, that was directed at you.

misty ruin
# tacit saddle

Adjust Environment Variables: Railway typically sets NODE_ENV=production automatically. In modern npm versions, this environment variable alone triggers the behavior of excluding devDependencies, making the explicit --production flag redundant and a source of the warning.

#

last patch which is simple make a startup command on railway and in that try to add --production in it which might overide this issue bcz your command will tell railway that yeah it is a production build

frank sage