# Google Cloud outage: what’s your plan when the cloud stops? URL: https://casterconsulting.hu/en/cloud-outage-plan-b/ Published: 2026-10-22 Updated: 2026-10-22 Author: Caster Consulting > Tuesday's Google Cloud outage didn't prove that the cloud is unreliable, it proved that most companies don't know what they depend on. An SMB doesn't need its own data centre. It needs a list of its critical services and a plan B for the day one of them goes down. **Short answer: Tuesday's outage wasn't really about the cloud. It was about dependency.** In a London data centre the backup generators failed to start, and a few hours later people couldn't pay by card with a bank most of them didn't know ran on Google. An SMB can't prevent its providers from ever going down. What it can decide is whether the whole company stops with them. ## What happened? According to Google Cloud's own incident report, multiple services in the europe-west2 (London) region had problems between 06:00 and 12:10 UTC on 20 October. The trigger was a power interruption at the data centre hosting the europe-west2-b zone. The controllers that start the backup generators faulted, so the generators never came on. The facility ran on batteries for as long as it could, and then the servers, network and storage in that zone lost power. The more interesting part comes next. Google says the impact didn't stay inside one zone. Traffic shifted to the other two zones, and dependencies between zones caused slowdowns and errors there too. The list of affected products runs to about fifty items, from virtual machines to databases to load balancing. Google's preliminary report on 22 October says the full root-cause analysis is still under way. The consequences showed up in everyday life. The Irish Times reports that Revolut customers in Ireland and the UK couldn't log in, transfer money, pay by card or withdraw cash for several hours. Revolut said customers' money and data were safe. Reuters, in a story republished by The Star, says Lloyds and HSBC were also named among those affected. Hungarian tech site HWSW adds the parcel company DPD, the NHS and Royal Mail to the list. Regulators moved quickly. According to Reuters, the European Banking Authority (EBA) contacted Google Cloud to understand the incident and its impact on EU financial firms, and the UK's Financial Conduct Authority is engaging with Google as well. Last year the EU designated 19 technology firms as critical third-party providers to the financial sector. This is the first serious test of that regime. ## What does this mean for an SMB? You probably don't have a Google Cloud account. You were still affected. That's the point. Your bank, your e-commerce platform, your accounting software, your invoicing tool and your CRM all run in someone's cloud, and you rarely find out whose. Tuesday showed that one power failure in that chain is enough to wipe out a company's morning. No attack required, and no bad intent. **A typical pattern is a company with one bank account and one way to pay.** If that's the day a supplier invoice falls due or the rent has to go out, "a few hours of disruption" stops being a technical problem. The same goes for an online shop: if the payment provider is down, the baskets stay full and the orders don't come in. **None of this makes the cloud worse than your own server.** Your own server room goes down too, it just doesn't publish an incident report. The difference is that with your own server you know where the risk sits. In the cloud you have to find out. As we wrote in our [article on moving to the cloud](/en/moving-to-the-cloud-smes/), responsibility is shared: the provider looks after the data centre, but what you do when something is unavailable is up to you. In financial services this is already written into the rules. Reuters notes that under both EU and UK rules, banks remain ultimately responsible for the services they provide, even when they rely on third-party providers. No such rule applies to an SMB. The logic still holds. ## What should you do now? 1. **List your critical services.** Which five systems can the company not work without for a single day (bank, email, invoicing, online shop, ERP)? Next to each one, write down the provider and who you call when it goes down. 2. **Have a second way to pay.** A second account or card with a different bank is cheap insurance for the day the first one doesn't work. The same applies to your online shop: if you can, keep a fallback payment method. 3. **Ask your key software providers** which cloud and how many zones their system runs in, what happens if a data centre goes offline, and what the contract commits them to for recovery. No answer is an answer too. 4. **Write a one-page plan B.** Who decides during an outage, how you tell customers and staff, and which minimal process you can run on paper or offline to get through the day. 5. **Keep your own copy of the most important data.** Customer list, open invoices, current orders: they shouldn't only be reachable inside the provider's cloud. If you'd like someone to map what your business depends on and where it makes sense to build in a fallback, take a look at our [infrastructure](/en/infrastructure/) service. ## Sources - [Google Cloud Service Health – Multiple services were experiencing issues in the europe-west2 region](https://status.cloud.google.com/incidents/Xs2XTHzAVC9YnafnPyzj) - [Reuters (The Star) – Europe's financial regulators query Google over cloud outage that hit banks](https://www.thestar.com.my/tech/tech-news/2026/10/21/europe039s-financial-regulators-query-google-over-cloud-outage-that-hit-banks) - [The Irish Times – Millions affected by Revolut disruption linked to Google Cloud outage](https://www.irishtimes.com/business/2026/10/20/revolut-services-hit-by-outage-including-in-ireland/) - [HWSW – Előszedték a Google-t a Revolutot is kifektető keddi üzemzavar miatt](https://www.hwsw.hu/hirek/71132/google-cloud-uzemzavar-bankszektor-europai-bankhatosag-vizsgalat.html)