This should be a warning to anyone running GCP. They suspend accounts left right and centre without even thinking about what they're doing. It seems like they use Gemini 3.1 Pro to run their production decisions.
TK has a history of absolutely destroying the culture of the place like in OCI and has done something similar in GCP from what I've heard. GCP and Google are completely different entities with how they work. Don't expect Google quality from the name. It's just like those old brands which now have cheap licensed products like Nokia (An exaggeration I know but not far from truth).
Not only that they are known to shut off their services randomly giving you like 6 months to migrate. They have lots of engineers not doing anything, so they put them on migrating internal users off those services, most of their clients don't. There was a brilliant article on this by an ex-GCP employee that I can't find right now.
Avoid GCP like plague if you are serious about your business.
And this is Railway, a big enough name to top the HN main page and presumably find someone from Google to intervene at some point. I would have zero recourse if it was some little product that I built.
Their account was restored in 10 / 19 minutes! It just took 4-6 hours to get everything fully healthy. I look forward to seeing the google response to this hopefully.
May 19, 22:10 UTC - Our automated monitoring detected API health check failures and paged our on-calls, who started investigating the issue.
May 19, 22:11 UTC - Dashboard returning 503 errors. Users unable to log in.
May 19, 22:19 UTC - Root cause identified: Google Cloud Platform has suspended Railway's production account.
May 19, 22:22 UTC - P0 ticket filed with Google Cloud. Railway's GCP account manager engaged directly.
May 19, 22:29 UTC - Incident declared.
May 19, 22:29 UTC - GCP account access restored. All compute instances remained stopped and persistent disks inaccessible.
100% agree, I've seen on Twitter and HN small players facing similar issues with no recourse and response from Google. I don't know what kind of place they are trying to build there.
They got TK to woo the enterprise customers who were forced to be hostage to OCI. But it seems they are still doing opposite of hostage here.
GCP was never known for their support and deprecation of services was always a huge risk. Its very sad because its actually a quality product. They should easily be the number 2 provider. Azure is extremely unreliable and their documentation is subpar. GCP being in 3rd place is more of their doing.
Yeah, sadly know that from being burned from one of their depreciations. In fact, 2-3. But you live and you learn. And it is better to learn from other's mistakes always.
Hasn't theGoog acted this way of quick to suspend accounts well before Gemini? I like to bash on LLMs as much as the next guy, but this seems very much like the memory of a gold fish. Or, you are just too young to remember pre-LLMs???
Haha, no. I know Google bans anyone randomly with usually no recourse in sight. I just wanted to take a dig at how bad their LLM is too while we were at it and thinks like Google themselves which is not surprising.
That seems to be the case. But as we see it backfires. Railway is very public, but we know at hacker news google has been doing this kind of thing for quite some time now.
Then it's a bad endorsement for Gemini 3.5 pro too. But jokes aside, I think they need a customer centric thinking instead of a self-centred one they seem to harbour even before TK joined (not everything can be blamed on him although it should be his responsibility now).
Google? Customer-centric? The closest thing to that is their cloud division buttering up some big name clients.
Other than that, Google prefers to act like "customers" are some kind of unfortunate rash they can't quite seem to get rid of, but would love to do so.
Meta is no different. I know a company that had their OAuth app on Meta rendered completely unusable just because one of their employees (a dev) had their personal Facebook account banned by Meta for no reason. They tried to escalate it multiple times but got nowhere, lol. Meta is even worse because accounts need to be 'personal'; if you have a Business Manager, the users added to it are all tied to their personal Meta/Facebook accounts. This is ludicrous.
Meta and Google B2B are both horrible. Their ad account bans are constant, and they have no real escalation process to get help. These companies are monopolies that should treat businesses more seriously, especially in these situations.
Yeah, people loose their business because a kid is logged in on their iPad, gets their google account suspended, and google knows it's the same household as the parent, and everything gets shut down
Are you honestly saying that a kid should not use their parent's email address as a recovery option? Seems like that would be the natural way to do it.
I don’t know about you, but I have a family account that we use as an email recovery for kids.
Adults have multiple emails so they won’t have to share it.
If something takes out the family email account, that’s fine. The only thing going there regularly are school notices, contractor receipts and recovery emails.
More businesses need to hear this message. Google has proven time and time again they cannot be trusted as a service provider, exactly because of this problem.
There is a history going back many years of Google suspending or terminating accounts with no explanation, often having to backtrack after users published their frustration and the incident went viral.
Google has always acted as if they have no obligations whatsoever to their paying customers.
They have not explained WHY their account was suspended. That's the most important part, imo. Cloud Providers don't suspend entire accounts for no reason.
> "Cloud Providers don't suspend entire accounts for no reason."
Maybe I'm getting old but here[1] is a HN comment from 17 years ago complaining about Google banning accounts "by mistake" and having no recourse but to post on HN and hope Matt Cutts sees it and helps, and saying "there are literally 1000s of such stories for many years all over the blogoshphere and forums" which is something I remember from HN of years ago.
In May 2024, Google Cloud Platform (GCP) accidentally deleted the private cloud account and all backups belonging to UniSuper, an Australian pension fund managing over $125 billion.
If it were something out of Railways hands, I think they would say something like "We have not yet identified the reason for the suspension, and are awaiting a response from Google".
At any company doing Enterprise work, you don't cut off someone for non payment without Account Manager doing multiple phone calls to whoever you have contact information for, emailing everyone listed on the account and whoever opened a support ticket and maybe even putting a banner in the panel with "ACCOUNT OVERDUE, CALL US TO SORT IT OUT!"
Generally it takes 30 days past due and complete no contact for anyone before suspension.
> Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action. This action extended to many accounts within Google Cloud. As this was a platform-wide action, there was no proactive outreach to individual customers prior to the restriction.
Never could. Google might block your entire company because one of your workers did something nasty on their personal account, and their ban hammer is mighty and blocks all related accounts to the Nth degree
I've been wondering if I should be interrogating my friends before allowing them access to my wifi. "Have you or any of your family members ever been banned by Google?"
In practice, Google has earned the way my priors are ready to believe it's 100% their fault with mighty and sustained effort. Or lack thereof, depending on your point of view.
> Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action. This action extended to many accounts within Google Cloud. As this was a platform-wide action, there was no proactive outreach to individual customers prior to the restriction.
Railway don't have a great reputation for building scalable systems (effects of vibe coding?). It's worth waiting for Google's response before jumping to conclusions. They can move to Azure/AWS/own datacenter, but there's a good chance this will repeat in a few months.
Sure, if this was one off isolated incident people would have agreed with you. But it's not. Even Google personal accounts have been used to ban their other ones including ones spending thousands of dollars on ads or GCP or any other paid google service, which is ridiculous.
Their reputation is fine, and their uptake is due in part to their handling of scaling.
If you're picking them instead of the underlying cloud provider, but you want all the knows and dials the underlying provider has, you've made the wrong choice.
Then let me be the not day-one account to say Railway is utterly bearing some responsibility here.
"However, in this ring, there was still a hard dependency on workload discoverability being tied to the network control plane API that was hosted on the machines running in Google Cloud."
They've gotta be joking me that they deliberately left something so critical under the control of any other entity than themselves. That demonstrates a lack of critical planning and a lack looking at their configuration from a first-principles approach.
There is always responsibility with Railway, that's given. But also taking into account how many big websites went down when AWS was down, building critical redundancy at such large scale is not cheap, and not many companies do it. Same as security theatre, we have redundancy theatre because they needed to sell the CLOUD.
This exact thing happened to me when I ran https://www.fatherly.com/ circa 2017. Google just shut down our account without notice. We were spending like $10k/month. It also locked us out of our premium support account, so we couldn't even get anyone there to notice that they'd locked us out.
After about 8 hours, a random Google support tech said it was because we were mining bitcoin, which was laughably untrue. We had CPU usage graphs and logs for the whole time and there was no spike. At around 12 hours, they turned it back on, said it was "misconfiguration of our abuse detection" and gave us like $100 in credit.
Absurd. Say what you will about AWS, they would never do that to a customer without a rep reaching out to you first. I have not trusted GCP since.
Google thinks everything should be replaced with automation.
Remember knowledge cards? Prior to the LLM AI revolution, they had an extraordinarily crappy AI system digest the entire internet to figure out the wrong facts about stuff and then present it to users as solid truth, with no human review and no way to report inaccuracies.
They just don't care. If the task requires a person to look at a thing and tell if it's right, they only do that for like 5 examples and then train a classifier, then deploy said classifier without thinking twice because "at internet scale" or whatever crap.
Google is the epitome of expecting happy path results to always be the end result. I could absolutely see someone writing this knowledge card system, but then realizing how much work it would be to edit it with some PM not wanting to say the project was a failure and needing serious amounts of human effort to correct and just releasing it as is. Gotta earn those KPIs for that next promotion, and then it's someone else's problem!
The report at this point is pretty much just a timeline of what happened. No explanation of why, no accusations, no blame. A PR piece, to Railway's customers, reassuring them that "we're not ignoring this."
Now the lawyers are huddling. IMO there won't be a lot more said publicly by either side, at least until any threat of lawsuits for damages is settled.
I don't think you're typically told why for these things, and it's mostly automated from what I can tell. The automated systems make mistakes but more importantly they're completely opaque. Nobody, not even Google, knows how they work exactly.
Google knows why there is no human oversight: because that is expensive (both in terms of the labor doing review and the ongoing fraud likely happening while the human review happens).
Really? This isn't the first time their automation took down a big customer (UniSuper in 2024) by accident. In that case the automation actually deleted the resources and GCP had to recover them.
Who is the "You" in "you haven't addressed the root cause"?
If you are asking Railway to spend effort doing this rather than simply moving away from GCP, I'm not sure why they would unless they want to sue GCP to recover damages to brand and long term customer retention.
The moment GCP shut off without any forewarning, its done deal, no need to ask any further questions.
"Railway owns our vendor choices, and we ultimately own this one. Your customers don't care whether the failure was Google or Railway; they see your product. Your uptime is our responsibility, and we'll keep delivering on it."
Kudos to them for acknowledging it and not doing PR speak. It shows it was an architectural failure from their part of trusting GCP, and they are working to fix it. Should they have seen it coming? Yes. But better late than never.
Railway has not had the best month in the tech press have they? And in both cases it was an automated process belonging to some other party that put them there, damaging their reputation.
I was going to talk to our google rep about their killing the Gemini cli but this is way more concerning.
Building on someone else's platform is always gonna be a risky move, and building a platform on top of someone else's platform is even riskier.
My company used to use a hosting provider that was basically AWS plus some extra guarantees. We just finished migrating onto regular AWS because they now offer what we need directly.
But...AWS is a platform too, no? Seems like you're in the same category of risk you just moved to a more well-known name. Granted, Amazon is the most reliable even if they have their own quirks.
In the case of them giving AI admin credentials to delete their production database, and it deleted their production database: that's on them. They were the only ones who put the admin account credentials into their AI.
Then they took no personal responsibility. That definitely damaged their reputation. Here, they are taking at least some responsibility. Props to them on improving.
Also, GCP does indeed have serious reliability issues, and Google does indeed have serious customer support issues.
EDIT: It has been brought to my attention below that the first 2 paragraphs are misattributed, and were not Railway, but rather a customer of theirs. Sorry, Railway!
Did Railway give admin credentials to delete their production database? My memory of the incident is that a customer of Railways used an AI tool to delete their production database, and then blamed Railway for it. The customer was the one who put their own account credentials into their own AI, not Railway
Unfortunately we had to make emergency migration off to Azure yesterday due to this. Thankfully our DB was not hosted on Railway and we were back up in a couple hours.
As much as we loved the simplicity they provided us, there's just been too many mishaps and shortcomings for us to continue running a B2B enterprise app on their infrastructure.
What were your reasons for going with railway in the first place? I'm not super familiar with them, but did you choose them for unique offerings, or essentially just VMs? If unique offerings, how rough was the migration out?
Question: for a smaller SaaS tool, or even internal product. If a team doesn't want to manage AWS or another IaaS provider, what are the best alternatives for the following
I have read plenty of snark about them on HN, but I found their product incredibly useful, well-designed, and easy to work with. If I was building a new startup from scratch, I'd definitely be giving them a look.
I'm sure there are plenty of the like 1,000 AWS products that DO has no viable competitor for, but for what they do offer, they're great.
If you are unable to use IaaS directly. You need to accept that your service might be down.
Even if you use AWS and the like, if you aren't building your app with redundancy across multiple AZs, then you'll have some downtime occasionally.
And even if you do build redundancy with multiple AZ, some services might fail anyway as AWS is not entirely isolated. So you might have downtimes.
So just accept downtimes and use the best tool for you (unless they are really bad, like GitHub level bad).
If you cannot accept any downtime, you'll have to spend millions of dollars and months of work to have the confidence to expect no downtime. Something like Netflix's chaos monkey and infrastructure would be enough.
The advantage of going with AWS is that when us-east-1 goes down, half the internet goes down so you don't have to defend why you had a service outage.
Yup. I don't know enough people at giant companies to know how many actually do this though. Not just talking having 2 AZs, I'm talking about ability in a DR scenario to fail over, within 5-10 minutes, to a different cloud provider, e.g. AWS → Hetzner, or GCP → Azure.
My gut feeling is that the number of significant applications that have this capability can probably be counted on two hands. Especially since a lot of the largest footprints of software stacks running in the cloud belong to Google and Microsoft, who I'm pretty sure do not replicate their services into someone else's cloud.
Before the cloud it was commonplace to have redundant data centres from two or more colocation provider companies. Similarly, Internet uplink diversity was commonplace.
An intermediary can provide value but there’s also a risk so I’d consider why you don’t want to use AWS, GCP, etc. directly. All of the major cloud providers have services which are only slightly harder than what Railway does but allow you to grow into more advanced things as your needs expand without adding a third-party who controls your features, security, and availability.
As an example, I note that GCP responded within 7 minutes according to their timeline. If you’d been using Cloud Run, that would have reduced downtime by over 7 hours — and there’s a good chance that you never would have gone down in the first place if the unknown trigger event was related to other customer activity or something odd Railway did.
There’s also a complexity factor: note how much complex infrastructure they mentioned having to fix that you wouldn’t need for your own account. That code does useful things, I’m sure, but it’s also a lot of moving parts which a hosting provider needs and you don’t – this outage took everyone down, whereas individual AWS or bare metal users would’ve otherwise been unaffected. There isn’t a global optimum which is the same for everyone but I think developers are prone to wildly over-estimating how much time they save by removing a couple of deployment steps relative to the direct costs and the less obvious costs of working within someone else’s environment.
This entire thread illustrates why you don’t want Google in any critical part of your business. AWS, sure. Azure? Maybe. I’m not familiar with Azure, but if I have to pick one, it’s AWS.
I love Fly, but their docs are.. tough. They've had multiple iterations of the control plane API, and it's very hard to do things the "correct" way with conflicting official docs.
Depending on exactly what you're building, all of these things sounds like one VPS. A bit of maintenance/security burden managing the machine if you're not used to it but as the others have said: Next.js can be selfhosted, unless you need the serverless/edge stuff; then I would go to Cloudflare Workers.
But really any service (or even on-site hosting) can have downtime, if that's not acceptable then I suppose building/using a tool that can be distributed between multiple hosts located in different geographical areas is the best option.
Haven't used railway but my understanding is they are something similar to Heroku. Fly.io has been pretty great for tiny projects in that niche.
For Vercel if your nextjs site can be compiled statically you could probably throw it up on almost anything. We've self hosted before which is pretty straightforward but you lose a lot of the image optimization stuff unless you go deep into setting up open next.
I love how everyone in Silicon Valley acts like Microsoft doesn’t exist.
Azure!
It’s the enterprise cloud with enterprise support. They won’t randomly pull the plug on your account, unlike companies that have a wildly different cultural background:
Google - ad tech (you’re the product)
Amazon - shop front (you’re a comptetitor)
Oracle - lawyers (you’re a future lawsuit for license extortion)
> May 19, 22:10 UTC - Our automated monitoring detected API health check failures and paged our on-calls, who started investigating the issue.
> At 22:20 UTC on May 19, Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action.
If the timestamps are accurate, what was causing the errors 10 minutes before the account was suspended?
The simplest explanation is just that one or the other of these timestamps is wrong, which wouldn't be a big deal. But if the timestamps aren't known with certainty, it seems very odd to include them in the writeup as though they are certain, even though they are very obviously inconsistent with each other.
> If the timestamps are accurate, what was causing the errors 10 minutes before the account was suspended?
Assuming the timestamps are accurate, Google probably started terminating resources while the account was not "suspended" and only completed that after all resources were disabled.
Or the account started doing something nefarious (assuming one of their customers as root cause, not railway itself) that started causing real problems and Google shut it down.
The problem with not having the data is that it’s easy to make assumptions.
The absence of any explanation for the suspension does seem intentional. If it were me that's one of the first things I would've asked so that I could make sure it doesn't happen again.
The 22:20 timestamp from the body of the post is wrong. The timeline section (where the 22:10 timestamp came from) is consistent with itself, and also contains:
> May 19, 22:19 UTC - Root cause identified: Google Cloud Platform has suspended Railway's production account.
They couldn't have identified the root cause before it happened.
That 10 minutes is likely very normal. Possibly...
* A Google employee messes up a setting (like one of the previous incidents) triggers something that looks like a suspension is warranted and it takes 10 minutes to flow through the process to suspend.
* A Railway customer does something corrupt, or seemingly corrupt, Google's system starts limiting access and take 10 minutes to decide it should be a suspension.
These are even more likely if there is a person in the loop to approve, who obvious did not dig deep enough to see that they should not have done so.
I've read all the threads and their main page and I still don't really understand what this service is. Is this like a commercial alternative to Gerrit? What do people use this for?
Here is my source code
Run it on the cloud for me
I do not care how
In this case it looks like they also bundle together a bunch of the other services you would need to get code onto the platform, monitor it once it’s there and so on
Oh I see, so they manage the server hosting and application server configuration, optimization and all that jazz. Almost like one step away from managed hosting. Makes sense now, thankyou!
What drives Google to apply these actions so completely and immediately, versus a more deliberate approach, with notification and delay before action, manual review for paying customers, or a warning to resolve within X hours/days? Once or twice could be errors or bad implementation, but these can't explain away the pattern.
It would seem that Google's counsel has deemed that whenever _____ is detected, the company must immediately and completely sever the business relationship. What is that driving concern? Is it sanctions enforcement? CSAM? Something else?
I work in a security adjacent role and I know we have had a few incidents that involved Railway networks lately. Could be something to that, I don't know
The problem is scale. Google uses automation and doesn't have the people to review the actions of that automation. I never worked at Google but this is the most obvious explanation from watching these things happen for years and years.
Please, someone that worked at Google, please comment.
"Your customers don't care whether the failure was Google or Railway; they see your product. Your uptime is our responsibility, and we'll keep delivering on it." - Thanks Claude!
Even if it ultimately turns out to be "Google's fault" (as this report seems to be saying), Railway say they own the incident but make no apology here.
The RCA and preventive measures was a pleasant read. I got a lot of respect for companies putting a lot of effort into incident reports like these. Makes them appear very professional rather than just blaming the cloud provider outright.
Honestly, i have been wanting to suggest to my leaders that we should go to on-prem for primary, and use cloud only as extra for peak traffic and/or failover, etc...but, the culture where i'm at is so bought into cloud as if it solves all problems...and then, in the next breath they all ask me to drastically reduce cloud costs and ensure 100% uptime at all times 24/7/365 (1005 uptime without complexity and without any added costs!).
It really is amazing that there is not some level at which "human review" becomes mandatory. Customers of that size already have dedicated account rep contacts.
I can't believe Kurian has not put his foot down about this. Adverse action against accounts over $X ARR absolutely must have review by revenue-carrying people before the action is taken.
I think this is just the default endgame of large corporates which suck up large quantities of customers. They are a race to the bottom and you end up with service by footgun. My own company is responsible for doing this in our sector. Literally every technology decision favours automation over verification because it's cheaper to say sorry than do it right.
Amazon played AWS from day 1 as if they were the runner-up (and in a sense they were), and while it does look like it's day 2 there, they are not letting the momentum down
Microsoft might have technical warts but commercially they are strong and Azure is a lot of times bundled with other services and you know you can get someone on the phone if needed
A spot in the Top 3, and is neither Microslop nor AWS.
At least that's my understanding from discussing with people praising GCP.
Let's say you want a big cloud provider, but you don't want Azure because of Microslop's old and recent history, and you don't want AWS because it's the default cloud provider.
You're left with GCP. And many people are stuck in the 00's, and still believe Google is the cool kid crushing the boring old corporations.
Honestly they really are starting to look that way. Total opinionated Walled Garden that's against an open and thriving ecosystem. Unlike Microsoft the technology is not yet garbage but I hope this isn't where they're going to end up
It sure is heading towards being garbage, though. Search is actively being degraded in favor of a barely functioning AI, and I'm sure it's not going to stop there. Seems like it was inevitable once ad/finance people got ahold of the company.
Why would you use an infrastructure provider on top of another infrastructure provider? It adds cost and risk, it's always going to be a leaky abstraction, and it's not hard to learn how to use GCP or AWS correctly - especially with agents.
I don’t understand why Google still has TK helming GCP when its obviously not achieved the kind of success it should. Google infra is some of the best in the world yet GCP is meh. It continues to underperform and seems content to be a distant 3rd behind AWS and Azure.
Google has a culture problem. This is not something that can change easily nor will it change when it’s not recognized as being an issue within their organization.
Between my peer c-suites, the conversation is that GCP cannot even be in the consideration set until such a time as a several-year period has elapsed without this kind of incident.
Now given the logic that you can't be dependent on any one service to run your SaaS, how does Railway convince its customers to run their SaaS on a single service?
I've been getting serious, recently, about moving all my workloads to equipment that I control in datacenters with which I have professional relationships. It's less expensive, easier, and this kind of nonsense doesn't happen. These cloud providers need to step back and observe how terrible they've made these products. Footguns everywhere, pricing that is impossible to forecast or reason about, broken APIs, and automated self destruction. Then you have third-party providers sitting on top of them, adding another layer of each antifeature. Crazy.
> These cloud providers need to step back and observe how terrible they've made these products.
They don't, because the allure of effortless scaling is hard to resist: everyone thinks of themselves as the next tech unicorn. And if you actually become an unicorn, you're already too dependent on AWS / Azure / GCP to easily move somewhere else. At best, your strategy is to become "multi-cloud".
That effortlessness is a fantasy. That's illustrated right here in this write-up by how complicated their system is.
>Railway’s network is a mesh ring, built up of high availability fiber interconnects between Metal <> GCP <> AWS. However, in this ring, there was still a hard dependency on workload discoverability being tied to the network control plane API that was hosted on the machines running in Google Cloud
"We had all of our workers set up in an open office layout to to make sure everyone could talk to each other without a single point of failure. But last night the boss got too drunk and didn't come in, so everyone spent the day scrolling tiktok."
It's really surprising how much cheaper colo becomes if you have an even vaguely predictable workload. And you don't have to be a major customer, either -- the data centers will happily sell you single U's or a couple U's, even on a monthly basis if you ask, making it perfectly viable for startups or advanced personal projects.
> ...These cloud providers need to step back and observe how terrible they've made these products...
I doubt that will happen because none of them want to stop the money-making machine they have! And, if your thought after my comment is that all us techies are making a fuss, so the cloud providers and businesses using them will hear our cries and trigger a backlash...? I doubt that to...because some senior business leaders that i see are bent on listening more to management consultants as opposed to abalance of folks including their own internal experts...but, alas, maybe i'm just having too cynical a day today. :-)
The thing that's nice about physical datacenters with people is that they often have to physically walk over to disconnect you - it's not as easy as some automated system doing an AI.
> Your customers don't care whether the failure was Google or Railway; they see your product.
Refreshing. So tired of businesses blaming their vendors. Oh it wasn't us spamming you text messages and emails, it was Shopify. Oh, our delivery guarantee said 2 days and it's been a week? That's not us, it's UPS.
I don't care. I didn't pay UPS or Shopify. I paid you.
> Railway’s production account into a suspended status incorrectly, as part of an automated action.
Be it individuals or companies, this time is the best time to ditch all dependence on anything clouds or SaaS since all are using automated AI, more and more of these incidents will occur.
So, what was the reason for the account suspension. Why did it happen? I know Google can be a bit stupid with their automatons but I am bit skeptical here. There are sites more critical than Railway hosted on GCP.
This should be a warning to anyone running GCP. They suspend accounts left right and centre without even thinking about what they're doing. It seems like they use Gemini 3.1 Pro to run their production decisions.
TK has a history of absolutely destroying the culture of the place like in OCI and has done something similar in GCP from what I've heard. GCP and Google are completely different entities with how they work. Don't expect Google quality from the name. It's just like those old brands which now have cheap licensed products like Nokia (An exaggeration I know but not far from truth).
Not only that they are known to shut off their services randomly giving you like 6 months to migrate. They have lots of engineers not doing anything, so they put them on migrating internal users off those services, most of their clients don't. There was a brilliant article on this by an ex-GCP employee that I can't find right now.
Avoid GCP like plague if you are serious about your business.
Edit: Gemini (unironically) found the article on this, a very good read: https://steve-yegge.medium.com/dear-google-cloud-your-deprec...
And this is Railway, a big enough name to top the HN main page and presumably find someone from Google to intervene at some point. I would have zero recourse if it was some little product that I built.
Their account was restored in 10 / 19 minutes! It just took 4-6 hours to get everything fully healthy. I look forward to seeing the google response to this hopefully.
May 19, 22:10 UTC - Our automated monitoring detected API health check failures and paged our on-calls, who started investigating the issue. May 19, 22:11 UTC - Dashboard returning 503 errors. Users unable to log in. May 19, 22:19 UTC - Root cause identified: Google Cloud Platform has suspended Railway's production account. May 19, 22:22 UTC - P0 ticket filed with Google Cloud. Railway's GCP account manager engaged directly. May 19, 22:29 UTC - Incident declared. May 19, 22:29 UTC - GCP account access restored. All compute instances remained stopped and persistent disks inaccessible.
100% agree, I've seen on Twitter and HN small players facing similar issues with no recourse and response from Google. I don't know what kind of place they are trying to build there.
They got TK to woo the enterprise customers who were forced to be hostage to OCI. But it seems they are still doing opposite of hostage here.
This is the bigger point of all of this. Scary.
> Don't expect Google quality from the name.
It sounds exactly like what I have experienced in terms of Google quality over the decades.
GCP was never known for their support and deprecation of services was always a huge risk. Its very sad because its actually a quality product. They should easily be the number 2 provider. Azure is extremely unreliable and their documentation is subpar. GCP being in 3rd place is more of their doing.
i wouldnt really call it a product without support, let alone a good product.
its a nicely design hobby, that somebody could make a good product out of, by following the same abstractions
All google products work like this. Should never be used for anything critical.
Yeah, sadly know that from being burned from one of their depreciations. In fact, 2-3. But you live and you learn. And it is better to learn from other's mistakes always.
> Don't expect Google quality
Google has an extremely poor reputation. Why are you thinking differently to that?
Hasn't theGoog acted this way of quick to suspend accounts well before Gemini? I like to bash on LLMs as much as the next guy, but this seems very much like the memory of a gold fish. Or, you are just too young to remember pre-LLMs???
Haha, no. I know Google bans anyone randomly with usually no recourse in sight. I just wanted to take a dig at how bad their LLM is too while we were at it and thinks like Google themselves which is not surprising.
They probably do use similar tech to make some of these decisions, though. And they always have done that as well.
You missed the humor part and focused on the tech part, it seems.
What/who is TK? What is OCI?
Thomas Kurian (GCP ceo)
That Google poached from Oracle Cloud, fwiw.
OCI is oracle cloud infrastructure.
This feels like google applying the same anti-spam mindset everywhere: detect risk, ban first, ask questions later.
That seems to be the case. But as we see it backfires. Railway is very public, but we know at hacker news google has been doing this kind of thing for quite some time now.
yep, last year they deleted UniSuper private cloud too
It's pretty stupid that big customers like Railway are not somehow protected from this.
I think all customers should be protected by at least one CSR doing a quick approve before banning the account.
> It seems like they use Gemini 3.1 Pro to run their production decisions.
They said they are already using Gemini 3.5 Pro internally.
Then it's a bad endorsement for Gemini 3.5 pro too. But jokes aside, I think they need a customer centric thinking instead of a self-centred one they seem to harbour even before TK joined (not everything can be blamed on him although it should be his responsibility now).
Google? Customer-centric? The closest thing to that is their cloud division buttering up some big name clients.
Other than that, Google prefers to act like "customers" are some kind of unfortunate rash they can't quite seem to get rid of, but would love to do so.
Yup, updated with the article I mentioned by Steve Yegge. Still holds true today.
"Finally, we are in planning to remove Google Cloud services from our data plane’s hot path, and keeping them only for secondary/failover."
That's pretty clear. Google can no longer be trusted as a B2B service provider.
Meta is no different. I know a company that had their OAuth app on Meta rendered completely unusable just because one of their employees (a dev) had their personal Facebook account banned by Meta for no reason. They tried to escalate it multiple times but got nowhere, lol. Meta is even worse because accounts need to be 'personal'; if you have a Business Manager, the users added to it are all tied to their personal Meta/Facebook accounts. This is ludicrous.
To me, building any business with dependencies on Meta is just a bad business plan.
Meta and Google B2B are both horrible. Their ad account bans are constant, and they have no real escalation process to get help. These companies are monopolies that should treat businesses more seriously, especially in these situations.
Yeah, people loose their business because a kid is logged in on their iPad, gets their google account suspended, and google knows it's the same household as the parent, and everything gets shut down
Can't find this now but google did at least once disable company's accounts after dev got their account suspended.
And as we know from the recent Gemini ban wave, you can get suspended just because.
> google knows it's the same household as the parent,
Nearly all these linkages are due to people sharing recovery email addresses and phone numbers. Don't do that.
Are you honestly saying that a kid should not use their parent's email address as a recovery option? Seems like that would be the natural way to do it.
I don’t know about you, but I have a family account that we use as an email recovery for kids.
Adults have multiple emails so they won’t have to share it.
If something takes out the family email account, that’s fine. The only thing going there regularly are school notices, contractor receipts and recovery emails.
It’s almost impossible not to any more. This is victim blaming at this point.
Everyone needs a defensible root of trust, this goes all the way down to the registrar you use for your domain.
More businesses need to hear this message. Google has proven time and time again they cannot be trusted as a service provider, exactly because of this problem.
They trust them enough to still give them money, just goes to show how entrenched big tech is and why they need to be broken up into dozens of pieces.
In the meantime start by migrating away from them for anything serious.
There is a history going back many years of Google suspending or terminating accounts with no explanation, often having to backtrack after users published their frustration and the incident went viral.
Google has always acted as if they have no obligations whatsoever to their paying customers.
Somebody else's computer
They have not explained WHY their account was suspended. That's the most important part, imo. Cloud Providers don't suspend entire accounts for no reason.
> "Cloud Providers don't suspend entire accounts for no reason."
Maybe I'm getting old but here[1] is a HN comment from 17 years ago complaining about Google banning accounts "by mistake" and having no recourse but to post on HN and hope Matt Cutts sees it and helps, and saying "there are literally 1000s of such stories for many years all over the blogoshphere and forums" which is something I remember from HN of years ago.
[1] https://news.ycombinator.com/item?id=791004
And unfortunately nothing has changed since then regarding this.
The cloud provider in question - GCP - who also deleted a 125 billion dollar company's entire account on accident?
What company?
In May 2024, Google Cloud Platform (GCP) accidentally deleted the private cloud account and all backups belonging to UniSuper, an Australian pension fund managing over $125 billion.
I think that stretches what it means for a company to "be" a 125b company, but that is still awful.
> Cloud Providers don't suspend entire accounts for no reason.
You're joking, right?
LOL, did you woke up from the hibernation?
This is Google we're talking about. This absolutely happened many times in the past and will happen again.
Google has suspended entire accounts countless times for absolutely no reason.
Unfortunately the cloud providers also rarely if ever tell you the reason.
Not defending them. but wouldn't it be a legal nightmare if they did?
My guess would be the credit card expired....
If it were something out of Railways hands, I think they would say something like "We have not yet identified the reason for the suspension, and are awaiting a response from Google".
At any company doing Enterprise work, you don't cut off someone for non payment without Account Manager doing multiple phone calls to whoever you have contact information for, emailing everyone listed on the account and whoever opened a support ticket and maybe even putting a banner in the panel with "ACCOUNT OVERDUE, CALL US TO SORT IT OUT!"
Generally it takes 30 days past due and complete no contact for anyone before suspension.
No one pays $2m invoices with credit cards.
FTA:
> Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action. This action extended to many accounts within Google Cloud. As this was a platform-wide action, there was no proactive outreach to individual customers prior to the restriction.
This might be 100% of what google told them.
Never could. Google might block your entire company because one of your workers did something nasty on their personal account, and their ban hammer is mighty and blocks all related accounts to the Nth degree
I've been wondering if I should be interrogating my friends before allowing them access to my wifi. "Have you or any of your family members ever been banned by Google?"
Hasn't every cloud provider had issues? Is the enshitification of servces really isolated to Google, or are we all doomed.
Banning accounts for no stated reason is kind of a Google speciality. They have a long well documented history of this sort of thing.
Railway has an overwhelming incentive to pin the blame on Google. This report doesn't answer why Google suspended Railway's account.
I'd wait for more details before adjudicating.
In principle, I agree with you.
In practice, Google has earned the way my priors are ready to believe it's 100% their fault with mighty and sustained effort. Or lack thereof, depending on your point of view.
That would be approximately 6365262822 time Google suspended someone for no good reason.
So no, Google doesn't get the benefit of the doubt.
They said it was automated and affected a bunch of other customers, which gives at least some hint.
And in general Google lost any immediate benefit of the doubt status many years ago. Many such stories.
To quote the article:
> Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action. This action extended to many accounts within Google Cloud. As this was a platform-wide action, there was no proactive outreach to individual customers prior to the restriction.
I'm not sure that's the lesson to learn from this outage. Hell Google resolved the problem in 7 minutes which is as good as you could hope for.
The resulting action should be you have proper disaster recovery, failover, etc.
Not sure I would trust these folks if this is the conclusion they are coming to from this experience. Any cloud provider can/will do this to you.
Google restored access but did not resolve the problem. VM’s were still shut down.
Google resolved the problem in seven minutes for a billion dollar company. Good luck if you are a nobody.
The best you could hope for is that if there is something fishy with your account, you are contacted by Google to address it.
Railway don't have a great reputation for building scalable systems (effects of vibe coding?). It's worth waiting for Google's response before jumping to conclusions. They can move to Azure/AWS/own datacenter, but there's a good chance this will repeat in a few months.
Sure, if this was one off isolated incident people would have agreed with you. But it's not. Even Google personal accounts have been used to ban their other ones including ones spending thousands of dollars on ads or GCP or any other paid google service, which is ridiculous.
Their reputation is fine, and their uptake is due in part to their handling of scaling.
If you're picking them instead of the underlying cloud provider, but you want all the knows and dials the underlying provider has, you've made the wrong choice.
There is always one bootlicker, fresh 1 day account no less.
Then let me be the not day-one account to say Railway is utterly bearing some responsibility here.
"However, in this ring, there was still a hard dependency on workload discoverability being tied to the network control plane API that was hosted on the machines running in Google Cloud."
They've gotta be joking me that they deliberately left something so critical under the control of any other entity than themselves. That demonstrates a lack of critical planning and a lack looking at their configuration from a first-principles approach.
There is always responsibility with Railway, that's given. But also taking into account how many big websites went down when AWS was down, building critical redundancy at such large scale is not cheap, and not many companies do it. Same as security theatre, we have redundancy theatre because they needed to sell the CLOUD.
The interesting and yet-to-be-explained part is why google flagged the account?
Put all the timestamps you want in the post mortem about what you observed, but you haven't addressed the root cause.
The "this doesn't make sense" part of the story likely has a real explanation that nobody wants to reveal yet.
This exact thing happened to me when I ran https://www.fatherly.com/ circa 2017. Google just shut down our account without notice. We were spending like $10k/month. It also locked us out of our premium support account, so we couldn't even get anyone there to notice that they'd locked us out.
After about 8 hours, a random Google support tech said it was because we were mining bitcoin, which was laughably untrue. We had CPU usage graphs and logs for the whole time and there was no spike. At around 12 hours, they turned it back on, said it was "misconfiguration of our abuse detection" and gave us like $100 in credit.
Absurd. Say what you will about AWS, they would never do that to a customer without a rep reaching out to you first. I have not trusted GCP since.
Google thinks everything should be replaced with automation.
Remember knowledge cards? Prior to the LLM AI revolution, they had an extraordinarily crappy AI system digest the entire internet to figure out the wrong facts about stuff and then present it to users as solid truth, with no human review and no way to report inaccuracies.
They just don't care. If the task requires a person to look at a thing and tell if it's right, they only do that for like 5 examples and then train a classifier, then deploy said classifier without thinking twice because "at internet scale" or whatever crap.
Google is the epitome of expecting happy path results to always be the end result. I could absolutely see someone writing this knowledge card system, but then realizing how much work it would be to edit it with some PM not wanting to say the project was a failure and needing serious amounts of human effort to correct and just releasing it as is. Gotta earn those KPIs for that next promotion, and then it's someone else's problem!
Shouldn't Google answer this if they are unhappy with this incident report? Are we even sure that Railway knows?
I seriously doubt Railway knows. That's the MO for Google and others, suspend account without explanation.
They can't - that would violate the privacy rights of their customer.
They need to tell Railway and Railway needs to tell us, or Railway can tell us that Google is refusing to tell them.
Either way, we need to hear about this from Railway.
The report at this point is pretty much just a timeline of what happened. No explanation of why, no accusations, no blame. A PR piece, to Railway's customers, reassuring them that "we're not ignoring this."
Now the lawyers are huddling. IMO there won't be a lot more said publicly by either side, at least until any threat of lawsuits for damages is settled.
the Railway PM doesn't say they weren't told. It just sort of glosses over this. I would be interested to know if they were told (or not).
I don't think you're typically told why for these things, and it's mostly automated from what I can tell. The automated systems make mistakes but more importantly they're completely opaque. Nobody, not even Google, knows how they work exactly.
Google should know why a human accepted the automated suggestion, or if and why there wasn't any human oversight in the first place.
Google knows and wants that there is no oversight. Don't do business with any big tech, if you don't want this kind of incidents.
Google knows why there is no human oversight: because that is expensive (both in terms of the labor doing review and the ongoing fraud likely happening while the human review happens).
For big accounts, like railway, zero chance this was a handsoff fully automated ban
Really? This isn't the first time their automation took down a big customer (UniSuper in 2024) by accident. In that case the automation actually deleted the resources and GCP had to recover them.
That assumes a competent org. If this were aws, I fully believe that. At gcp it's entirely plausible.
Who is the "You" in "you haven't addressed the root cause"? If you are asking Railway to spend effort doing this rather than simply moving away from GCP, I'm not sure why they would unless they want to sue GCP to recover damages to brand and long term customer retention.
The moment GCP shut off without any forewarning, its done deal, no need to ask any further questions.
That‘s the point where Google tells you they won’t tell you the exact reason because of security reasons
Exactly this, which is the problem with all modern accounts. No person to talk to so you can understand what happened and maybe fix it.
They most definitely have a person to talk to. They're not the largest Google Cloud user by far, but they are large enough to have human account reps.
And those reps might not be told what the reason is.
They also don't want to tell you because then they have to put rules and cannot ban people arbitrarily.
Giving reasons is putting accountability on Google and they don't want that.
This isn’t the first time Google Cloud has seriously messed with a customer’s account: https://cloud.google.com/blog/products/infrastructure/detail...
"Railway owns our vendor choices, and we ultimately own this one. Your customers don't care whether the failure was Google or Railway; they see your product. Your uptime is our responsibility, and we'll keep delivering on it."
Kudos to them for acknowledging it and not doing PR speak. It shows it was an architectural failure from their part of trusting GCP, and they are working to fix it. Should they have seen it coming? Yes. But better late than never.
it sounds to me like templated text from the UVic ESS office somewhere:P
Railway has not had the best month in the tech press have they? And in both cases it was an automated process belonging to some other party that put them there, damaging their reputation.
I was going to talk to our google rep about their killing the Gemini cli but this is way more concerning.
Building on someone else's platform is always gonna be a risky move, and building a platform on top of someone else's platform is even riskier.
My company used to use a hosting provider that was basically AWS plus some extra guarantees. We just finished migrating onto regular AWS because they now offer what we need directly.
But...AWS is a platform too, no? Seems like you're in the same category of risk you just moved to a more well-known name. Granted, Amazon is the most reliable even if they have their own quirks.
Each critical dependency you stack multiplies your risk. Now you have to worry about Railway AND Google causing business-damaging outages.
In the case of them giving AI admin credentials to delete their production database, and it deleted their production database: that's on them. They were the only ones who put the admin account credentials into their AI.
Then they took no personal responsibility. That definitely damaged their reputation. Here, they are taking at least some responsibility. Props to them on improving.
Also, GCP does indeed have serious reliability issues, and Google does indeed have serious customer support issues.
EDIT: It has been brought to my attention below that the first 2 paragraphs are misattributed, and were not Railway, but rather a customer of theirs. Sorry, Railway!
Did Railway give admin credentials to delete their production database? My memory of the incident is that a customer of Railways used an AI tool to delete their production database, and then blamed Railway for it. The customer was the one who put their own account credentials into their own AI, not Railway
You are totally right, my mistake.
Unfortunately we had to make emergency migration off to Azure yesterday due to this. Thankfully our DB was not hosted on Railway and we were back up in a couple hours.
As much as we loved the simplicity they provided us, there's just been too many mishaps and shortcomings for us to continue running a B2B enterprise app on their infrastructure.
Sad day :(
What were your reasons for going with railway in the first place? I'm not super familiar with them, but did you choose them for unique offerings, or essentially just VMs? If unique offerings, how rough was the migration out?
Azure suspended your account as well?
I think they meant that they migrated off of railway TO azure as opposed to FROM azure
Question: for a smaller SaaS tool, or even internal product. If a team doesn't want to manage AWS or another IaaS provider, what are the best alternatives for the following
1.) Vercel - having a bad month
2.) Supabase - having a bad month
3.) Railway - now having a bad month
DigitalOcean. Seriously. They have been around a long long time and built a lot of the core infrastructure you rely on every day (e.g. Ceph).
I've used DigitalOcean for personal projects for over a decade, no major issues so I definitely recommend!
I;ve had my share of VPS & Managed DB outages at DO, so they are also not faultless.
I've been with DO since checks mailbox 2014. Honestly never experienced an unannounced outage.
Yeah overall they are ok. I think 3 times managed db and one or twice a vps just dead. No issues in a year or so.
They were always hardware failures, took about 45-120min. Not the end of the world, but also not fun getting lot of client complaints.
Not if, but when. No one is faultless. Chasing after 100% is a fool's errand.
I have read plenty of snark about them on HN, but I found their product incredibly useful, well-designed, and easy to work with. If I was building a new startup from scratch, I'd definitely be giving them a look.
I'm sure there are plenty of the like 1,000 AWS products that DO has no viable competitor for, but for what they do offer, they're great.
I've had nothing but good experiences with them and their docs and tutorials are excellent.
Yes, I use DO with Hatchbox. It is a perfect combo. Been using for more and more projects.
If you are unable to use IaaS directly. You need to accept that your service might be down.
Even if you use AWS and the like, if you aren't building your app with redundancy across multiple AZs, then you'll have some downtime occasionally.
And even if you do build redundancy with multiple AZ, some services might fail anyway as AWS is not entirely isolated. So you might have downtimes.
So just accept downtimes and use the best tool for you (unless they are really bad, like GitHub level bad). If you cannot accept any downtime, you'll have to spend millions of dollars and months of work to have the confidence to expect no downtime. Something like Netflix's chaos monkey and infrastructure would be enough.
The advantage of going with AWS is that when us-east-1 goes down, half the internet goes down so you don't have to defend why you had a service outage.
I just blame AWS for all outages whether that’s true or not.
I think the message here is that you can't trust any single cloud provider. You at least need two with full operational capability.
Yup. I don't know enough people at giant companies to know how many actually do this though. Not just talking having 2 AZs, I'm talking about ability in a DR scenario to fail over, within 5-10 minutes, to a different cloud provider, e.g. AWS → Hetzner, or GCP → Azure.
My gut feeling is that the number of significant applications that have this capability can probably be counted on two hands. Especially since a lot of the largest footprints of software stacks running in the cloud belong to Google and Microsoft, who I'm pretty sure do not replicate their services into someone else's cloud.
Before the cloud it was commonplace to have redundant data centres from two or more colocation provider companies. Similarly, Internet uplink diversity was commonplace.
An intermediary can provide value but there’s also a risk so I’d consider why you don’t want to use AWS, GCP, etc. directly. All of the major cloud providers have services which are only slightly harder than what Railway does but allow you to grow into more advanced things as your needs expand without adding a third-party who controls your features, security, and availability.
As an example, I note that GCP responded within 7 minutes according to their timeline. If you’d been using Cloud Run, that would have reduced downtime by over 7 hours — and there’s a good chance that you never would have gone down in the first place if the unknown trigger event was related to other customer activity or something odd Railway did.
There’s also a complexity factor: note how much complex infrastructure they mentioned having to fix that you wouldn’t need for your own account. That code does useful things, I’m sure, but it’s also a lot of moving parts which a hosting provider needs and you don’t – this outage took everyone down, whereas individual AWS or bare metal users would’ve otherwise been unaffected. There isn’t a global optimum which is the same for everyone but I think developers are prone to wildly over-estimating how much time they save by removing a couple of deployment steps relative to the direct costs and the less obvious costs of working within someone else’s environment.
This entire thread illustrates why you don’t want Google in any critical part of your business. AWS, sure. Azure? Maybe. I’m not familiar with Azure, but if I have to pick one, it’s AWS.
Ramnode had always worked well for my projects.
Fly, Render, and even Heroku still are all better choices then working with Railway I think
I love Fly, but their docs are.. tough. They've had multiple iterations of the control plane API, and it's very hard to do things the "correct" way with conflicting official docs.
Why does nobody consider that you can buy a baremetal box or even a VPS and that will get you very far without paying a metered fee
Hatchbox + Digital Ocean is an unbeatable combo and provides Railway-like automation with self-owned infra.
Depending on exactly what you're building, all of these things sounds like one VPS. A bit of maintenance/security burden managing the machine if you're not used to it but as the others have said: Next.js can be selfhosted, unless you need the serverless/edge stuff; then I would go to Cloudflare Workers.
Maybe a VPS? Simple to manage and way cheaper.
But really any service (or even on-site hosting) can have downtime, if that's not acceptable then I suppose building/using a tool that can be distributed between multiple hosts located in different geographical areas is the best option.
Haven't used railway but my understanding is they are something similar to Heroku. Fly.io has been pretty great for tiny projects in that niche.
For Vercel if your nextjs site can be compiled statically you could probably throw it up on almost anything. We've self hosted before which is pretty straightforward but you lose a lot of the image optimization stuff unless you go deep into setting up open next.
Fly.io (AFAIK) still has a relatively good track record?
Hetzner (or any VM provider) + Dokku works best.
I love how everyone in Silicon Valley acts like Microsoft doesn’t exist.
Azure!
It’s the enterprise cloud with enterprise support. They won’t randomly pull the plug on your account, unlike companies that have a wildly different cultural background:
Google - ad tech (you’re the product)
Amazon - shop front (you’re a comptetitor)
Oracle - lawyers (you’re a future lawsuit for license extortion)
Etc…
Shameless self plug but check out: https://specific.dev (especially if you use coding agents)
No code lock-in through SDKs and built on top of AWS with great DX for both developer and coding agents
> May 19, 22:10 UTC - Our automated monitoring detected API health check failures and paged our on-calls, who started investigating the issue.
> At 22:20 UTC on May 19, Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action.
If the timestamps are accurate, what was causing the errors 10 minutes before the account was suspended?
The simplest explanation is just that one or the other of these timestamps is wrong, which wouldn't be a big deal. But if the timestamps aren't known with certainty, it seems very odd to include them in the writeup as though they are certain, even though they are very obviously inconsistent with each other.
> If the timestamps are accurate, what was causing the errors 10 minutes before the account was suspended?
Assuming the timestamps are accurate, Google probably started terminating resources while the account was not "suspended" and only completed that after all resources were disabled.
Or the account started doing something nefarious (assuming one of their customers as root cause, not railway itself) that started causing real problems and Google shut it down.
The problem with not having the data is that it’s easy to make assumptions.
The absence of any explanation for the suspension does seem intentional. If it were me that's one of the first things I would've asked so that I could make sure it doesn't happen again.
The 22:20 timestamp from the body of the post is wrong. The timeline section (where the 22:10 timestamp came from) is consistent with itself, and also contains:
> May 19, 22:19 UTC - Root cause identified: Google Cloud Platform has suspended Railway's production account.
They couldn't have identified the root cause before it happened.
That 10 minutes is likely very normal. Possibly...
* A Google employee messes up a setting (like one of the previous incidents) triggers something that looks like a suspension is warranted and it takes 10 minutes to flow through the process to suspend.
* A Railway customer does something corrupt, or seemingly corrupt, Google's system starts limiting access and take 10 minutes to decide it should be a suspension.
These are even more likely if there is a person in the loop to approve, who obvious did not dig deep enough to see that they should not have done so.
> As a side effect, Terms-of-service acceptance records were also reset, prompting users to re-accept on their next visit to the dashboard.
Don't get me wrong- the rest of this mess falls pretty clearly on Google Cloud, but this one feels like something Railway did to themselves.
I've read all the threads and their main page and I still don't really understand what this service is. Is this like a commercial alternative to Gerrit? What do people use this for?
I'm not a developer, just curious what this is.
The category is “Platform as a Service”
Alternative to Fly or Heroku
Here is my source code Run it on the cloud for me I do not care how
In this case it looks like they also bundle together a bunch of the other services you would need to get code onto the platform, monitor it once it’s there and so on
Oh I see, so they manage the server hosting and application server configuration, optimization and all that jazz. Almost like one step away from managed hosting. Makes sense now, thankyou!
What drives Google to apply these actions so completely and immediately, versus a more deliberate approach, with notification and delay before action, manual review for paying customers, or a warning to resolve within X hours/days? Once or twice could be errors or bad implementation, but these can't explain away the pattern.
It would seem that Google's counsel has deemed that whenever _____ is detected, the company must immediately and completely sever the business relationship. What is that driving concern? Is it sanctions enforcement? CSAM? Something else?
It could be automated action based on abuse reports. TONS of spam comes from Railway associated networks.
I work in a security adjacent role and I know we have had a few incidents that involved Railway networks lately. Could be something to that, I don't know
The problem is scale. Google uses automation and doesn't have the people to review the actions of that automation. I never worked at Google but this is the most obvious explanation from watching these things happen for years and years.
Please, someone that worked at Google, please comment.
Sadly, My Railway project is still having issues 24 hours later. Already started emergency migration away from Railway backend :(
It's highly unlikely that GCP banning their account without telling them is true, but GCP is probably not going to go public with the real reason.
I’m 100% willing to believe it based on Google’s track record.
Duplicate of:
https://news.ycombinator.com/item?id=48201484
Looks like it was endorsed by dang: https://news.ycombinator.com/item?id=48210941
"Your customers don't care whether the failure was Google or Railway; they see your product. Your uptime is our responsibility, and we'll keep delivering on it." - Thanks Claude!
Unfortunately we've also had a litany of problems with our GCP deployment and chose to remove them completely as a service provider.
Had similar experience with GCP. Terminated VMs six times, and responded zero times.
Even if it ultimately turns out to be "Google's fault" (as this report seems to be saying), Railway say they own the incident but make no apology here.
The RCA and preventive measures was a pleasant read. I got a lot of respect for companies putting a lot of effort into incident reports like these. Makes them appear very professional rather than just blaming the cloud provider outright.
back to on-prem
Honestly, i have been wanting to suggest to my leaders that we should go to on-prem for primary, and use cloud only as extra for peak traffic and/or failover, etc...but, the culture where i'm at is so bought into cloud as if it solves all problems...and then, in the next breath they all ask me to drastically reduce cloud costs and ensure 100% uptime at all times 24/7/365 (1005 uptime without complexity and without any added costs!).
I will definitely not be signing up on GCP because of this.
It's reassuring to know they will ban a million dollar enterprise customer just like they will ban your GMail of 20 years.
It really is amazing that there is not some level at which "human review" becomes mandatory. Customers of that size already have dedicated account rep contacts.
I can't believe Kurian has not put his foot down about this. Adverse action against accounts over $X ARR absolutely must have review by revenue-carrying people before the action is taken.
Google, the new Microsoft!
I think this is just the default endgame of large corporates which suck up large quantities of customers. They are a race to the bottom and you end up with service by footgun. My own company is responsible for doing this in our sector. Literally every technology decision favours automation over verification because it's cheaper to say sorry than do it right.
I’m no Microsoft fan, but they are pretty good at long term support of enterprises. Way, way better than Google from what I can tell.
Amazon played AWS from day 1 as if they were the runner-up (and in a sense they were), and while it does look like it's day 2 there, they are not letting the momentum down
Microsoft might have technical warts but commercially they are strong and Azure is a lot of times bundled with other services and you know you can get someone on the phone if needed
Google has... ?
At least that's my understanding from discussing with people praising GCP.
Let's say you want a big cloud provider, but you don't want Azure because of Microslop's old and recent history, and you don't want AWS because it's the default cloud provider.
You're left with GCP. And many people are stuck in the 00's, and still believe Google is the cool kid crushing the boring old corporations.
> Google has... ?
former Oracle salespeople
Honestly they really are starting to look that way. Total opinionated Walled Garden that's against an open and thriving ecosystem. Unlike Microsoft the technology is not yet garbage but I hope this isn't where they're going to end up
It sure is heading towards being garbage, though. Search is actively being degraded in favor of a barely functioning AI, and I'm sure it's not going to stop there. Seems like it was inevitable once ad/finance people got ahold of the company.
Why would you use an infrastructure provider on top of another infrastructure provider? It adds cost and risk, it's always going to be a leaky abstraction, and it's not hard to learn how to use GCP or AWS correctly - especially with agents.
What intermediate is involved?
19 minutes from detection to getting the google account restored is pretty awesome honestly.
They forgot to get reimbursement for downtime. A free month of GCP is better than nothing.
I don’t understand why Google still has TK helming GCP when its obviously not achieved the kind of success it should. Google infra is some of the best in the world yet GCP is meh. It continues to underperform and seems content to be a distant 3rd behind AWS and Azure.
Flagged by some AI automation.
Google has a culture problem. This is not something that can change easily nor will it change when it’s not recognized as being an issue within their organization.
Between my peer c-suites, the conversation is that GCP cannot even be in the consideration set until such a time as a several-year period has elapsed without this kind of incident.
TK TK TK TK. Welcome to the jungle.
Now given the logic that you can't be dependent on any one service to run your SaaS, how does Railway convince its customers to run their SaaS on a single service?
I've been getting serious, recently, about moving all my workloads to equipment that I control in datacenters with which I have professional relationships. It's less expensive, easier, and this kind of nonsense doesn't happen. These cloud providers need to step back and observe how terrible they've made these products. Footguns everywhere, pricing that is impossible to forecast or reason about, broken APIs, and automated self destruction. Then you have third-party providers sitting on top of them, adding another layer of each antifeature. Crazy.
> These cloud providers need to step back and observe how terrible they've made these products.
They don't, because the allure of effortless scaling is hard to resist: everyone thinks of themselves as the next tech unicorn. And if you actually become an unicorn, you're already too dependent on AWS / Azure / GCP to easily move somewhere else. At best, your strategy is to become "multi-cloud".
That effortlessness is a fantasy. That's illustrated right here in this write-up by how complicated their system is.
>Railway’s network is a mesh ring, built up of high availability fiber interconnects between Metal <> GCP <> AWS. However, in this ring, there was still a hard dependency on workload discoverability being tied to the network control plane API that was hosted on the machines running in Google Cloud
What the hell is even that?
"We had all of our workers set up in an open office layout to to make sure everyone could talk to each other without a single point of failure. But last night the boss got too drunk and didn't come in, so everyone spent the day scrolling tiktok."
It's really surprising how much cheaper colo becomes if you have an even vaguely predictable workload. And you don't have to be a major customer, either -- the data centers will happily sell you single U's or a couple U's, even on a monthly basis if you ask, making it perfectly viable for startups or advanced personal projects.
> ...These cloud providers need to step back and observe how terrible they've made these products...
I doubt that will happen because none of them want to stop the money-making machine they have! And, if your thought after my comment is that all us techies are making a fuss, so the cloud providers and businesses using them will hear our cries and trigger a backlash...? I doubt that to...because some senior business leaders that i see are bent on listening more to management consultants as opposed to abalance of folks including their own internal experts...but, alas, maybe i'm just having too cynical a day today. :-)
The thing that's nice about physical datacenters with people is that they often have to physically walk over to disconnect you - it's not as easy as some automated system doing an AI.
And if they do, you can walk over there too and ask a human why in person. (Or just call the NOC)
Does this qualify for a list entry on killedbygoogle.com ?
> Your customers don't care whether the failure was Google or Railway; they see your product.
Refreshing. So tired of businesses blaming their vendors. Oh it wasn't us spamming you text messages and emails, it was Shopify. Oh, our delivery guarantee said 2 days and it's been a week? That's not us, it's UPS.
I don't care. I didn't pay UPS or Shopify. I paid you.
>Google Cloud placed Railway’s production account into a suspended status incorrectly, as part of an automated action.
There is no justification given on why this action was incorrect. It's possible they actually did something wrong.
But not so wrong that they could get their account back in 10 minutes.
Yup...
Major infra provider -> has no backups/game plan if GCP goes down
> Railway’s production account into a suspended status incorrectly, as part of an automated action.
Be it individuals or companies, this time is the best time to ditch all dependence on anything clouds or SaaS since all are using automated AI, more and more of these incidents will occur.
So, what was the reason for the account suspension. Why did it happen? I know Google can be a bit stupid with their automatons but I am bit skeptical here. There are sites more critical than Railway hosted on GCP.
Related discussion during the incident:
https://news.ycombinator.com/item?id=48201484
Perfect reminder that it's time to use Google Takeout while I still can.
tldr: AI suspended an almost a billion dollar startup account.