c7acf011-372d-4a33-92fc-8e2b9fe89610

@Brody can u help me pls?

have you checked your Primary logs? looks like its complaining about a missing json file.

my guess would be that you have incorrect config on the service left over from when it deployed from a github repo, such as a start command.

The service just stopped this morning, it was all doing fine, i haven't made any changes in the enviroment, or in the repo

your activity logs show you making a lot of changes to the primary service

Yeah, because it just stopped this morning and i`m tryin to just redeploy it to see if it works?

Since i didn't made any changes

Hey Brody, totally understand — but just to clarify:

When I said I didn’t make changes, I meant I didn’t touch the codebase or environment variables before the service went down this morning.

All the actions in the logs were me trying to restart or redeploy the service after it suddenly stopped, hoping it would fix the issue.
If there’s an old config or leftover setting from the initial GitHub deployment, that might be it, but I haven’t modified anything manually in the repo recently.

Would appreciate any pointers on where exactly that misconfig might be.

Thanks!

Starting Container

Initializing n8n process

There was an error initializing DB

connect ECONNREFUSED ::1:5432

User settings loaded from: /home/node/.n8n/config

Last session crashed

There was an error initializing DB

User settings loaded from: /home/node/.n8n/config

Last session crashed

let me try some things, you have likely misconfigured a bunch of stuff

please go hands off the computer

Bro what?

im trying to fix what has been missconfigured, thats a little hard when you are applying my changes

Hey Brody,

Let me set the record straight:

I didn’t change or misconfigure anything. The service stopped completely on its own, without any intervention from my side. I wasn’t even using n8n at the time — I found out it was down because another independent instance running on DigitalOcean alerted me that the webhook from Railway wasn't responding.

Everything you’re seeing in the logs is me trying to restart or redeploy after the crash, just to get it working again.

Also worth noting: I’m not the only one reporting this issue. There are other users facing similar problems, so this clearly isn’t a local misconfig on my end — it’s more likely a platform-level issue or something introduced upstream.

I’d appreciate if we could focus on diagnosing the root cause instead of assuming user error.

Let me know how I can assist.

Also, I just saw your message asking me to stay hands off — no problem. I’ll leave it untouched from now on so you can troubleshoot it freely.

i'm sorry but the AI responses are not appreciated, im trying to help

So try to be more polite tho

Never saw a customer support like this lol

And yeah, of course i'm using AI to craft a response to you, as i noticed you didn't understand me before ¯_(ツ)_/¯

this is unfortunately user error, your primary service has been failing since 4 weeks ago, this is not a platform issue

The health checks were failing because of some custom nodes I added to the Dockerfile in the repo — they were meant to support MCPs before N8N had native tools for that.

But the main service itself was running fine. Like I mentioned, the app only stopped working this morning.

My landing page frontend actually alerts me whenever webhooks don’t respond.

it does that by pinging another N8N instance I have running on DigitalOcean.

N8N is now accessible again

in the future, please do not be so quick to blame the platform

Just a friendly tip for future situations: don’t assume user misconfiguration before confirming the facts.

It saves time and leads to smoother communication on both sides.

you can see all the changes i needed to make in the activity panel

in this case the assumptions are correct

Bro, seriously, show me the exact logs where I supposedly changed any environment variables. I’ll wait. 😅

What’s really bizarre is the arrogant stance you're taking, even after I’ve repeatedly made it clear that I didn’t touch any environment variables, or the Railway environment at all.

you had the private network disabled, that was the biggest issue i found.

im sorry you think im taking some kind of stance, i am simply trying to help while diverting the unnecessary blame away from the platform.

please confirm if you can log in to n8n

Alright, thanks for pointing that out.

I wasn’t aware the private network was disabled

That said, I genuinely didn’t change anything in the n8n service.

The only thing I deployed was a separate Facebook conversions API, in the same environment, but completely isolated from n8n, Redis, or Postgres

So I’d really like to understand.

How could that have affected the main service?

I’m asking because I want to avoid issues like this in the future, not to point fingers.

all communications are done via the private network, with that disabled neither the worker or the primary will be able to talk to the database

hence the connect ECONNREFUSED ::1:5432 error, thats the pg client defaulting to trying to find postgres at localhost because it cannot connect over the private network

Thanks for the clarification.

Just to add, my landing page sends data to both the conversions API and a webhook on n8n to store events and cookies.

Is there any chance that certain payloads or request types from the frontend could somehow trigger the private network to be disabled or misconfigured?

Just trying to understand what could’ve caused it.

All fine with N8N, just logged in

the only way to disable it would be to toggle it off from within one of the service settings.

Fine, thanks for your help, have a good work

no problem!

!s