Twitter Touts Joyent Amid More Outages
Here's a scenario: You're a fast-growing Web 2.0 service under fire for your shaky uptime. You need to reassure your devoted but increasingly grumpy users that you can stay online during the Super Bowl. So you post a notice touting the merits of your "trusty infrastructure provider" - which just suffered a 10-day outage on one of the services it hosts.
It doesn't make sense to me either. (UPDATE: Apparently it didn't to Twitter once they thought about it. See our update.)But the folks at Twitter seem to believe that the time was right for a post titled Happy Happy Joyent, which notes that Twitter tends to be mobbed during large shared events:
It makes us glad that our trusty infrastructure provider Joyent has us covered with extra capacity — for free! In fact, throughout our amazing growth, Twitter has relied on Joyent's highly scalable infrastructure. While we're busy building Twitter, Joyent is working tirelessly to bring in more RAM, more CPUs, more hardware, and more late night support — never charging us for bandwidth is worth an extra mention. The decision to choose Joyent at an early stage was one of the big decisions that contributed to Twitter's success.In fact, Twitter was offline for more than six days in 2007 and has already logged more than 11 hours of downtime in January. That includes more than six hours of outages in the past two days.
Incredibly, that blog item was posted the day after Twitter was down for two hours during the State of the Union speech. Just hours after the post appeared, Twitter was offline for three hours, apparently due to attempted Twittering during the Republican debate last night.
As many readers are aware, two online backup services operated by Joyent, Strongspace and Bingodisk, were offline for more than a week after Joyent struggled to recover from a file corruption problem on it storage system. The Strongspace outage extended to 10 days.
It's not clear whether Joyent's performance is playing any role in Twitter's downtime, or whether the problems lie elsewhere in Twitter's infrastructure. At this point, taking note of problems at either Twitter or Joyent almost feels like piling on. But I'm baffled as to why these two services would be choosing this moment as the time to publicly congratulate one another.
UPDATE: In a follow-up item, Twitter says last night's outage was planned maintenance. "But we did go far beyond our planned time window, which sucks. What sucks more is the series of outages (planned and not) we've had lately. We know this makes Twitter frustrating to use, and we want you to know, we hear (and share) your frustration and are working really, really hard to fix it."
The company says it "finished a major infrastructure project tonight, which we've been working on for months and that we think is going to help a lot."
If you enjoyed this post subscribe to our RSS feed
By Rich Miller 
January 31, 2008 | Permalink | >Get Posts By E-mail
RELATED ENTRIES
Major Outage at Fasthosts -
Apr 15, 2008
Ships Impounded in Middle East Cable Cuts -
Apr 13, 2008
Hillary Clinton's Site Plagued by Outages -
Apr 10, 2008