TWO service outages within the course of several hours rocked microblogging platform Twitter on Thursday, as users worldwide reported significant downtime and slow service across both Twitter's website and mobile applications.
Amid speculation that Twitter had been crippled by a hacker attack, the San Francisco-based company blamed the outage - one of its most severe episodes in recent months - on a "cascading bug" in one of its infrastructure components.
"One of the characteristics of such a bug is that it can have a significant impact on all users, worldwide, which was the case today," Mazen Rawashdeh, a Twitter vice-president of engineering, wrote in a blog post on Thursday afternoon, after normal service resumed.
"We are currently conducting a comprehensive review to ensure that we can avoid this chain of events in the future," he said.
Twitter's statements came after UgNazi - an emerging hacker outfit that recently gained publicity for breaking into Cloudflare CE Matthew Prince's personal Google e-mail account - claimed credit for the service disruption in an e-mail to Reuters, saying it had launched a distributed denial-of-service (DDoS) attack against Twitter because of the company's support for the Cyber Intelligence Sharing and Protection Act.
One security professional said the group probably used a DDoS-for-hire site to launch an attack against Twitter on Thursday, but downplayed the likelihood of the group being solely responsible for bringing down the social media network.
"It was mere coincidence," the security professional said. "The back end of Twitter is having issues, which is unrelated to the very small attack."
North American traffic levels for Twitter.com sharply plummeted on two occasions between 3.30pm GMT and 6pm GMT, according to data provided by network analytics company Sandvine.
The first outage lasted between 3.30pm GMT and 5pm GMT, data showed.
Twitter acknowledged the disruption in a blog post that was continually revised as the service resumed, only to fail for a second time.
Thursday's sustained outage leaves a fresh bruise on a service that had supposedly long ago shed its reputation for being unreliable.
As the service resumed on Thursday, its most dedicated users quickly hopped back on to crack jokes, express relief and complain about the interruption - and, indeed, the fact that during the outage they had nowhere to complain about the interruption.
Founded in 2006, Twitter was plagued in its early days by frequent outages as its servers struggled to handle the ever-rising volume of tweets generated worldwide, leaving frustrated users with its famous "fail whale" error screen.
In recent years, Twitter, which has been under great pressure to demonstrate financial viability, has also devoted considerable resources to improving its reliability, in an attempt to project itself as a mature, polished brand.
CEO Dick Costolo said this month that Twitter now has 140-million active monthly users who send 400-million tweets daily.
The company sounded an apologetic note on Thursday, as it conceded it had failed users who rely on the platform to connect with "heroes, causes, political movements".
"It's imperative that we remain available around the world," said Mr Rawashdeh, "and today we stumbled."