The day of the great blackout, nothing here went dark

On 28 April, at 12:33, the Iberian Peninsula lost power. It was not a neighbourhood outage or a local fault: the electricity systems of mainland Spain and continental Portugal suffered a total blackout, with isolated effects in south-west France. This is confirmed by ENTSO-E, the European association of electricity network operators, which has already launched the technical investigation; early reports point to a combination of factors: grid oscillations, voltage-control and reactive-power problems, rapid drops in generation and cascading disconnections.

That Monday, millions of people discovered just how many things depend on a plug socket. And thousands of businesses asked themselves, now unavoidably, a question that until then had been theoretical: what happens to my systems when the power goes out?

In our Barcelona data centre, the answer was measured in milliseconds. At 12:33, the moment mains power was lost, the UPS systems — uninterruptible power supplies: batteries that hold the load at the very instant of the cut, with no perceptible transition — took over the full power supply. The servers noticed nothing: not a restart, not a flicker.

Then the data centre’s diesel generators started up and took over, preserving the batteries’ autonomy and sustaining the power supply for as long as the outage lasted. And so it went, hour after hour, until the grid came back and everything switched over again just as uneventfully. Our servers did not shut down and the service did not go down. A real test — not a scheduled drill, nor an audit with the date announced in advance.

It was not luck: it was architecture. We operate in a professional data centre in Barcelona — Econocom Nexica’s — with dual power lines, dual UPS systems and diesel generators, supervised in person 24/7, 365 days a year. And on that foundation, every one of our servers has dual power supplies connected to two independent electrical branches — the A/B scheme: if a line, a distribution board or a UPS fails, the other branch keeps feeding the machine. Every link in that chain existed long before 28 April; that day, it simply did its job.

Just as important: external connectivity was not lost. The services hosted on our infrastructure remained reachable from every network that still had electricity — unaffected areas, other countries. Anyone who could connect kept working.

And here is the honest caveat, because this house does not sell fairy tales: many users in Spain could not work that day. Not because the hosted systems failed, but because the access networks — the office fibre, the mobile phone masts — need electricity too, and many were left without it. Anyone with no power and no coverage could not reach any service, ours included. The lesson is twofold: the infrastructure responded, and business continuity has to be thought through end to end, access included.

What does all this mean for your business? That continuity is not decided on the day of the blackout, but long before: when you choose where and how your systems live. It is the criterion behind our infrastructure — redundant power, redundant network, separate racks — and the security we operate it with. And if your business keeps systems in the office, a well-designed hybrid cloud serves exactly this purpose: what is critical lives where there are batteries and generators, and the office is just one more access point.

The diagram below sums up that day in two lines: the electricity grid’s, which cuts off at 12:33 — and the green one, the service’s, which never moved.