Amazon and Twitter both saw this week how much the world relies on their services.
First, Amazon suffered a patch of downtime following a power failure in a North Virginia datacenter --- leading to a number of high-profile sites falling with it --- and many jumped to Twitter to complain.
Ironically, it was Twitter's turn to stumble a few days later. But the microblogging site's recent downtime generated a lot more buzz than one might have expected. Twitter goes down all the time, right?
Its users have become accustomed to Twitter "just working." People don't notice when a service is working: they only notice its absence when it's not. Only when the site failed to load with the once-favorite error message of the web --- the "Fail Whale" --- did its users start to wonder what was wrong.
Twitter has, over the past year, quietly reached a level of maturity en par with that of Facebook or any other major web service.
It's grown up. It's no longer the startup it once was. It became fully-fledged Web service, amid the acne, the growing pains, and the fallings out with its parents.
In its early days, even Google was temperamental. It would work, it wouldn't work. It's hard to think of that now. Back then, there was less of a search monopoly and fewer websites anyway. We knew exactly where to get our content from with or without a search engine.
And if not, we had AltaVista or Yahoo as firm and reliable backups.
But should Google stumble for more than ten minutes, the Web would be abuzz with concern for the company's stability --- and no doubt its stock price.
To see the level of outpouring to other social media sites during the outage shows just how far Twitter has come in the past six months to a year. Downtime of more than five minutes would still go mostly unnoticed. We still expect a level of instability --- as though we're collectively suffering from a post-traumatic experience --- but nothing could prepare us for an outage on this scale.
Compared proportionally to a year ago, this week's outage was a major one.
Had Twitter been hacked? Was Twitter's office move to blame? Had someone tripped over the master power cable?
No. It was just a bug. A cascading bug to be precise.
Twitter's vice president of engineering Mazen Rawashdehexplained on the company blog that it has seen "at least 99.96 percent" reliability, more "often 99.99 percent" during the past six months.
In fact, it's had a pretty good patch of uptime since it redesigned and rebuilt its infrastructure from the ground up. The company has effectively discarded its "playful" Fail Whale because on the most part, it doesn't need it anymore.
Dear Twitter, in the face of criticism and users complaining over the "scandal" of a downed service, with poise and maturity you fared the outage well. Amid the focus on the downtime, you should instead be applauded for a year of strong reliability.
Welcome to the grown-up club.
Image credit: Twitter.
- Tweetless in Seattle, also New York, San Francisco, etc.
- ZDNet: The truth about the Twitter crash
- CNET: Twitter taken down by nasty 'cascaded' bug
- CBS News: Twitter site down for brief outage