Twitter datacentre migration hit by problems

Microblogging service Twitter's datacentre migration has been plagued by difficulties, due to structural problems at its new, bespoke datacentre in Utah.

The site, which announced on 21 March that it had completed a major datacentre migration, was forced to change datacentre plans, mid-flight, due to a botched datacentre, according to sources who spoke to Reuters.

"A new, custom-built facility in Utah meant to house computers that power the popular messaging service by the end of 2010 has been plagued with everything from leaky roofs to insufficient power capacity," Reuters said.

If the sources are true, Twitter may have alluded to the problems throughout its 'Great Migration' blogpost on 21 March. At the time, Twitter described a two stage migration, where Twitter services were moved out of its primary datacentre and into a second staging datacentre, before moving to a third datacentre as the service's "final nesting ground".

Once it had touched down in the third facility, it moved services out of its first and second datacentres. "This essentially required us to move much of Twitter two times," according to the post.

According to Reuters, Twitter moved from a co-location facility it had been renting from hosting provider NTT America, into its bespoke Utah facility, but due to the problems migrated to a third datacentre owned by co-location company Raging Wire.

Twitter still has the use of its Utah facility, as it has signed a four year leasing contract. But, according to a Twitter source who spoke to Reuters, "nobody wants to put any part of the site that's mission critical in Salt Lake City... that would be a huge battle royale inside the company if they were told they had to."