Financial information firm Bloomberg suffered a two-and-a-half hour systems outage on Friday 17 April, cutting off hundreds of thousands of users of the firm’s Bloomberg terminals. 

The system was caused by “a combination of hardware and software failures”, said a brief company statement. Bloomberg spokespeople ruled out a deliberate attack, and specifically denied a rumour that the failure had been caused by someone spilling a drink in a server room. 

Restarting the software

The fault started around 8.30am London time, and left around 300,000 traders round the world unable to access the Bloomberg teminals which feed them with information. 

The loss of the world’s leading trading and data platform meant around 200 million fewer shares were sold on the London stock exchange, and the UK government’s Debt Management Office (a Bloomberg customer) had to postpone selling around £3 billion of debt. The bonds were sold in the afternoon when service was restored.

Around mid-day, Bloomberg reported that “service has been fully restored,” adding that the hardware and software failures had “caused an excessive volume of network traffic.”

Machines were then overwhelmed and customers were disconnected. “We discovered the root cause quickly, isolated the faulty hardware, and restarted the software,” said the statement. “We are reviewing our multiple redundant systems, which failed to prevent this disruption.”