UPS failure during maintenance cuts power at Maidenhead data centre
May 2012Published 13 Oct 2026Responsible editor: Chief Editor
Get the next one by email
Free alerts on data centre incidents in the United Kingdom, and an email when EIR-2322 is updated. No account needed; unsubscribe in one click.
Get a free sample brief Three complete verified records and the incidents-per-GW table, by email within a minute.
By signing up you agree to the privacy policy.
Companies named in this entry: Pulsant
What happened
On the night of 29 May 2012, Pulsant suffered a power disruption at its Maidenhead data centre that cut customer servers and firewalls off from the internet. During planned maintenance, an uninterruptible power supply (UPS) failed after a device swap. The company said a component changed during routine maintenance, a certified new manufacturer part, failed as the UPS was returned to service. The electrical load was taken over and then dropped, and customer racks connected to the failed UPS in the Maidenhead 2 and 3 facilities lost power.
The UPS was bypassed a few minutes later and power was restored, but some hosting customers were still recovering the next day after forced hard shutdowns. Pulsant initially described the problem as a network issue on its status page before reporting power issues. The UPS was awaiting full testing before being returned to service.
Contributing factors
A component fitted during routine maintenance failed when the UPS was placed back into service. Pulsant said the part was certified and from a new manufacturer. Customers said the status page initially described the fault as a network problem, and that night staffing levels slowed recovery. Pulsant said six engineers were onsite when the power was lost.
Lessons learned
- Test UPS units fully after any component replacement before returning them to service under load.
- Verify that replacement parts from new manufacturers are compatible and tested, not only certified.
- Keep status-page reporting accurate and update it promptly when the cause is identified as power rather than network.
- Plan maintenance so that a failed UPS does not drop load to customer racks, and confirm bypass and transfer paths work before work begins.
- Ensure out-of-hours staffing and escalation for hosting customers are adequate during maintenance windows.
More lessons on outages, backup supply and restoration, maintenance, inspection and ageing assets, from every incident on the register.
Source
Published 2026-10-13. Written from public reporting. Descriptive, not investigative. See the data accuracy disclaimer.
