Servers crash after data centre overheats
It's getting hot in here.
By Tom Jowitt | Techworld | Published: 16:07, 01 June 2009
The spate of hot weather in the UK claimed a notable scalp after a London data centre experienced a cooling failure, which caused several servers to overheat and crash.
The Braham Street data centre, located in the City of London, and owned by Level 3 Communications, experienced a chiller failure on Sunday when one of the five units designed to cool the data centre failed.
"On 31 May, we experienced a problem with one of the five chillers in place at the site," said James Heard, President of the European markets group, in an official statement from Level 3.
"The faulty chiller is currently being repaired, and the other chillers continue to operate at the facility," he added. "Customers are being informed of the issue."
Techworld understands that the data centre itself continued to run as usual with no downtime, but that several servers within the data centre itself overheated and crashed.
Although the outside temperature on Sunday in that area peaked at a high of 23 degrees Celsius, the internal temperatures within the data centre soared to an estimated 50°C by 7pm in the evening. Typically, data centres aim to run at anywhere from 18°C to 25°C.
These soaring temperatures claimed at least one victim, bringing down servers belonging to the music service Last.fm, for five hours approximately.
Last.fm confirmed the problem via its Twitter feed. "Crikey, one of our data centers (sic) has overheated! We're fixing it as fast as we can, but the site will be down for a bit," it wrote on Sunday.
"Apologies for the downtime, our data centre appears to have landed on the sun. More updates soon," it tweeted later.
The timing of the server crash for Last.fm was particularly unfortunate, considering that just days before, it had boasted (again via Twitter) about the uptime of one of its servers (1,244 days or 3.4 years).
Last.fm did not respond at the time of writing.
