Cloud provider Railway says Google Cloud suspended its account without cause, resulting in an outage; in 2024, GCP deleted an Australian pension fund's account
PaaS platform Railway says Google temporarily suspended its account on Wednesday without cause, inducing a major outage.
The Register
Context & Ripple Effects
The related coverage shows this is not an isolated concern in Google Cloud’s history: in 2018, Google said it would adjust automated fraud-detection suspensions after a customer complaint, and in 2024 it attributed UniSuper’s week-long private-cloud outage to rare issues that included an account deletion.
For Railway, the incident exposes a dependency inherent in operating a PaaS on a hyperscale cloud: a provider-level account action can become a service-wide availability event for downstream users.
First-order effects
Railway and its customers face immediate disruption from the suspension, while Railway must restore service and explain how a Google Cloud account action propagated into its platform outage.
Google Cloud faces a fresh reliability and support-process question, particularly because related coverage already documents a severe account-deletion incident affecting UniSuper.
Second-order effects
Railway’s customers may reassess whether their applications need independent backups, export paths, or deployment options outside a single underlying cloud account.
Other PaaS operators and enterprise cloud customers gain another reason to scrutinize suspension controls, escalation channels, and account-recovery commitments when choosing or renewing cloud providers.
Third-order effects
If account-enforcement failures recur, cloud reliability will be judged not only by infrastructure uptime but also by the governance, reversibility, and human review behind provider account controls.
The pattern strengthens demand for architectures and contracts that reduce the blast radius of a hyperscaler’s administrative action, though the practical trade-off remains added operational complexity and cost.
The trend: Cloud buyers are increasingly treating account-control resilience—rather than compute and network uptime alone—as a core dimension of platform risk.
Google Cloud has blocked our account, making some Railway services unavailable. We have escalated this directly with Google. The Railway Platform team has since confirmed access to Google Cloud and is working on restoring access to all workloads. We have access to some of our
This wild. Two years ago, Google deleted a large customer's account: Unisuper, an Australian retirement savings fund. It seems they now blocked an even larger customer, cloud provider Railway. This kind of story you never hear with AWS, Azure, or even Oracle. GCP 👎
Google Cloud blocked cloud infra provider Railway's account, taking it offline. It's not the first time GCP did something unacceptable: in 2024, they deleted the infra of a $124B Australian fund. Not fully trusting GCP saved the fund's data. Full: https://blog.pragmaticengineer.c…
Some thoughts on the discourse around @Railway team's outage. I was a mid-level SE at @datadoghq ANZ in 2023 when the company had a major, long lasting outage (actually started while I was mid demo). I was in a sales role, and what was most interesting to me was the response from
This is just unbelievable! Railway even has their own datacenters, but the APIs and edge network is on Google Cloud so everyone using Railway is down because Google just nuked their account 🤯
railway and every railway service going down because Google Coud blocked them was NOT on my bingo card I have heard horror stories of people getting blocked by google cloud and response/support being slow but this is wild, esp with the amount of spend they must be doing
When you are on GCP, you NEED to have a backup on another cloud, because Google can delete everything. Unisuper did this, and avoided full data loss only thanks to this. I assume Railway will no longer rely on only GCP. They would be foolish to do so. https://newsletter.pragmatic…
I wonder how much damage this does to Google Cloud, let alone Railway. Has to be at least $1B for google long term. As someone who worked in gcp, I would describe the culture as the opposite of customer obsession. I have a few horror stories I'll keep to myself, but let's just
We have recovered our compute on Google Cloud, but services are unable to start because of ongoing networking issues on Google Cloud's side. We are engaged with Google Cloud support to resolve this and will post the next update as soon as we have more to share.
It appears Google Cloud has blocked our account, and so some services are unavailable We've escalated this to Google and will keep people posted. Deepest apologies.
Does stuff like this ever happen with AWS? I only ever hear about it with GCP. It feels like they took the same technology they use to ban people's personal Gmail accounts and just pointed it directly at enterprise cloud customers running real businesses
Railway services have fully recovered. Some workloads may still need a redeploy, we're automatically redeploying any we detect as unhealthy. If your service isn't responding correctly, please trigger a redeploy from the dashboard or CLI. We're sorry for the disruption. A