Then, wanting to deploy a backup instance: Failed to create TCP proxy, please try again

This is a fuc*ing clown show right now

Same here - we're experiencing internal networking issues with our database deployment, which had been running smoothly for the past two weeks. The database also fails to connect through the simple viewer in the dashboard. This has occurred several times now over the last few months.

A redeployment of the service resolved it on our end, but it would be great to understand what's actually happening on Railway's side. Worth noting that a simple restart of the service does not fix the problem - only a full redeploy does.

Experiencing the same issue. Redeployment as opposed to restart worked for me.

My clients are getting quite annoyed with all of the outages. Prior to this year, Railway has been pretty rock solid for me, but this year... not great...

having issues with PR environments as well

things going down all the time

Its down AGAIN - are you kidding me?

Currently in the process of switching to Hetzner. Of course Railway is down on the last day. Looking forward to never using it again. Been customer for years.

i have same problem

I'm thinking the same thing... Railway is down almost every day.

I got an email for an infra maintenance event for said instance. Maybe they did sth in the background, maybe not (who knows, they are not communicating anything)

The following steps worked to recover if they didn't help in the background:

- Volume detached

- Instance duplicated

- Deployment removed from original

- Volume attached back to original

- Redployment triggered of original

back for 5 minute and than still problem

We are looking into a full migration away from Railway. Railway is not production ready. Too many outages. Our most active slack channel is the incident channel from Railway.

I’m also looking for a place to migrate to another provider because lately it’s been impossible to work with Railway.

Do you have any ideas, or have you already found something good? If so, let me know.

Redeploying helps for now but after some time we run into the same networking issues which don't seem to recover automatically.

Hi everyone, first off, apologies for the delay here. We had a hardware failure on one of our EU West hosts and services were migrated off the failing host, but the recovery process caused some follow-on disruptions which is why some of you might have seen things break again. The full incident timeline is at https://status.railway.com/incident/cmmui0c7z012icp7ebcd1a3zv.

We hear the frustration regarding reliability. The frequency of incidents in 2026 has not met the bar we set for ourselves or that you expect from us. We're investing heavily in scaling and faster failure recovery. We are growing now at a rate of over 9,000+ new users a day, and we're working hard to keep improving the experience while we scale through that growth.

@haayhappen: the volume detach/reattach workaround you found is solid. Thank you for sharing that with everyone here.

If anyone is still experiencing issues, please open a new thread with your project and we'll investigate directly.

I am still have problems with postgrees and redis aplications

Confirming this is still an issue.

Experiencing this issue now.

Any update?

Hanging on building containers. Any one else?

Hi all, if you're still experiencing connectivity issues with your Postgres or Redis services, or builds hanging, please open a new thread with your project link so we can investigate your specific services directly.