Custom Salt Lake City facility being built to address technical problems resulting from the microblogging site's rapid growth.
In a move designed to address some of the technical issues that have plagued the fast-growing microblogging site, later this year Twitter plans to move its technical operations infrastructure into a new, custom-built data center in the Salt Lake City area.
The site has struggled to deal with ongoing growth due to an average addition of about 300,000 new users per day, said Jean-Paul Cozzatti, Twitter engineering programming manager, in a company blog. In addition, Twitter combated some much-publicized outages during popular global events, including the recent World Cup.
Currently, Twitter uses San Francisco-area data centers operated by NTT.
"Having dedicated data centers will give us more capacity to accommodate this growth in users and activity on Twitter," said Cozzatti. "Importantly, having our own data center will give us the flexibility to more quickly make adjustments as our infrastructure needs change."
The new data center will have a larger footprint in a building designed specifically to meet Twitter's power and cooling needs, and Twitter will have total control over network and systems configuration, he said. Twitter will use a mixed-vendor environment, using servers running open source operating systems and applications, wrote Cozzatti.
"Finally, Twitter's custom data center is built for high availability and redundancy in our network and systems infrastructure. This first Twitter managed data center is being designed with a multi-homed network solution for greater reliability and capacity," he said. "We will continue to work with NTT America to operate our current footprint, and plan to bring additional Twitter managed data centers online over the next 24 months."
On Monday, Twitter grappled with a fault in the database that stores Twitter user records that caused problems on both Twitter.com and its API.
Last month, Twitter had 99.17% uptime and was out for 5 hours and 43 minutes, compared with 99.88% uptime and a 52 minute outage in May 2010, according to Pingdom. As of Thursday morning, Twitter had 99.88% uptime with 35 minutes down for July 2010, Pingdom found.
"We frequently compare the tasks of scaling, maintaining, and tweaking Twitter to building a rocket in mid-flight," Cozzatti said.