Kudos to the folks at Twitter, who seem to have gotten their biggest weakness -- reliability problems -- mostly sorted out this summer.
In Twitter's monthly email update, cofounder Biz Stone notes that Twitter's August uptime was 99.88%, as measured by Web monitoring service Pingdom -- 54 minutes of downtime. And this month, things are even better: Pingdom says Twitter's been up 99.97% of September, with just one minute of downtime. Both are a vast improvement over May, when Twitter was up 97.1% of the time -- or more than 21 hours of downtime.
Twitter's servers are responding quicker, too: Pingdom says Twitter's average response time is 1800 milliseconds, down from 2500 milliseconds in May.
Why do Twitter's uptime and performance matter? It's one thing for a social networking site or a blogging service to go down for a few minutes or hours at a time -- usually that's no problem. But if Twitter is going to grow and thrive as a utility -- like the Internet or the phone system, which users and partners trust as an always-on service -- it just needs to work.
There's still parts of Twitter that need some work, like its IM service, and the Flash badge that once proudly powered our Community Twitter page. But so far, it looks like Twitter's biggest stability problems are behind them.
Caveat: These are all measurements of Twitter's Web site, not the back-end Twitter service, which includes text messaging and third-party access via an API. So while Twitter Web downtime and Twitter service downtime aren't exactly the same, they should be roughly equivalent.
Source: Pingdom
See Also:
Twitter, AMC, Wise Up, Restore "Mad Men" Accounts
Free VC Advice For Startups, Via Twitter
Learning To Live With Less: Constraints, Rules And Web Startups
