I’m having the same issue, and it’s seriously making me reconsider things.

My app is supposed to launch in one week, and right now I honestly don’t know whether I should stay on Railway or move elsewhere. There seem to be problems almost every day, and I still don’t understand the cause.

I’m not sure whether this is related to broader infrastructure issues, regional instability, or something else, but until now Railway had been solid for me.

At this point I’m trying to figure out whether I should:

scale my current setup,

upgrade resources,

or start looking for a more stable hosting provider.

Are others seeing the same problems?

And for those who use other hosting platforms, are they more stable right now?

O variantă mai scurtă și mai naturală pentru forum:

I’m having the same problem, and honestly it’s starting to worry me.

My app launches in a week, and I don’t know whether I should stay on Railway or move to another provider. Lately there seem to be issues almost every day, and I can’t figure out why.

Railway had been stable for me until now, so I’m trying to understand whether this is temporary, whether I need to upgrade my setup, or whether I should look for a more reliable server.

Is anyone else experiencing this?

And if you’re using another hosting provider, is it more stable?

Marius, cel mai bine e sa stai pe Railway. Recent ei si-au create propriile datacentere ceea ce e mult mai greu de lucrat decat daca ramaneau pe GCP sau AWS. Dar asta inseamna ca au mult mai multa flexibilitate in a avea grija de infrastructura. Nu cred ca maine, poimaine, o sa fie totul rezolvat, dar fiind cu Railway de la bun inceput, pot zice cu tarie ca au o echipa super buna si o sa rezolve problemele.

I've been on railway for a few years. Since they moved to managing their own data center equipment things have been a bit unstable. But there is a lot more flexibility in doing that. I personally believe that Railway has a super solid team and will get back to having things running smoothly soon.

Multumesc de raspuns, sper sa rezolve cat mai curand, si eu am incredere in ei ca ofera cel mai bun serviciu de pe piata mai ales pentru proiecte la inceput.

Hi andrewmurraydavid, i'd frankly love to hear your reasons for wanting to stay.

These issues have been happening quite a bit. At first I was quite forgiving as I know this is an early stage product and company, which is why i've only been using it for personal, smaller projects, but the way that they handle customer issues is just ridiculous.

First:

They don't offer any support unless you pay extra. I currently pay hefty bills (Much higher than what you'd pay on aws for ex), but apparently this isn't a reason for them to offer technical support after they systems go down. (Unless you pay extra).

Then when they do get back to your through "Community threads", they auto close them without giving you the ability to reply.

So if you're looking for something "Reliable", well: AWS's UI sucks, but at least there will be no down times, you can reach their customer support at all times, and you'll much pay less.

Hey Max, I'm sorry you're experiencing this.

I've use many PaaSs and Cloud providers (Netlify, AWS, GCP, Azure, etc). One of my customers requested the infra be on AWS and most allow to be have the freedom on engineering whatever infra I choose fits their complexity. The costs of Railway are amazing, as an example the AWS-strict customer pay over $1200/mo for a basic SaaS where my bill for all my customers (7+) and company infra with anything from low to high complexity (multiple dbs, micro services, etc.) is under $350 (this includes the pro subscription for me and my devs). So there's no way to beat them at cost.

I find the observability to be great, but if you want even more advanced observability, just deploy a sidecar with API access to push the API events (logs, metrics, etc.) to a 3rd party service.

They one-click-deploy works great, especially for teams that love testing out different stacks, as railpack is really good at auto-determining the run env.

Let's talk about downtime, because that's primarily what this thread is based on. Yes, ever since they moved off of cloud provides and only managing their own bare metal, things have been a bit... questionable. But I know (from having been with them for a long time) what they are capable of. I have experienced the times where they deployment issues they had where from the atrocious github connector (which I believe is one of the reasons github it dying; a whole 'nother story). That being said, I do tend to use a lot of the beta features they have public on the user settings page and I keep up to date with their emails, releases, etc. to better understand what bottle necks they are facing. No company (mine including) will share bottlenecks before they're solved or have a plan to solve them, but it's good to see that they are acknowledging things.

Not sure how much of this comes off as fan-boy or real-life information, but I hope it's helpful.

Agreed, the past 6 months have been rough, we are planning a migration to fly.io because of this. They seem to have less outages.