In your metrics tab --> network graph, I'm seeing a big spike in network traffic recently. Does this correspond to when you see it crashing?

Yes, but its still down until now

?

Here's what I think the issue is:

When you're trying to handle SIGTERM signals in Node.js for graceful shutdowns, but you're running your app through npm commands like npm start, your signal handlers won't work properly. What happens is that your Node.js process actually runs as a child of the npm process, so when the system sends a SIGTERM signal, both npm and your app receive it.

Can you try running with just node not within npm? So instead of having something like web: npm start in your Procfile, you'd change it to web: node index.js. This way, your Node.js application becomes the top-level process and can properly receive and handle SIGTERM signals with whatever grace period you've coded in.

Hey, thank you for your response, I dont have a procfile, i m using the default nixpack

How I'm supposed to do?

I tried sth on api-staging, could you take a look please ? thank you

I've just push in api-prod as well

the app is still down, we dont understand we dont have any errors

✅ The internal ticket npm sigterm error has been marked as completed.

.

Can you send the repository of this project?

.

I can't access your private repository.

Can you give me your github handle?

We're using typeorm with typescript, does postgres on railway have a connection pool manager system ?

we've just move our database to supabase, we will remove the postgres if supabase is working

?

I ended up finding the solution myself, it turns out that your PostgreSQL template doesn't handle connection pooling. I would have appreciated a bit more support, especially when asking straightforward questions, rather than receiving emoji responses

We were considering moving to an enterprise plan, but this kind of experience makes us question whether it's really worth it

Thank you, and have a good day.

Hi HyperSwapX team, apologies about that — we had moved this question to community support as it appeared to be application specific (and not a Railway platform level problem).

Just wanted to mention, as one of the perks of our enterprise plan, you'd get a direct line with our team over Slack and you can reach out to us at anytime. In moments when enterprise users have urgent matters they would like to escalate to us, they can directly page us (yup, anytime - even 3 AM our time, including weekends).

If you're still considering it, I'd be happy to get you connected with our solutions team that would be able to answer any questions you've got to get going on Railway. Once again, sorry for the experience.