A week ago today, the unthinkable happened. That's right, Gmail went down...for *two* hours. You would have thought, judging from the amount of chatter on Twitter that we were experiencing an epic attack, a horrible natural disaster, perhaps the end of life as we know it; but it was none of that. Just couldn't get our email for a couple of hours. I'm surprised the Obama administration didn't step in and declare a State of Emergency. It certainly seemed plausible based on the reaction to the outage.
My favorite tweet, which captured perfectly, the level of hysteria we were seeing on Twitter as people learned Gmail wasn't working, was this one:
@scalzi: MIN 30 OF GMAIL OUTAGE THE CITIES ARE IN FLAMES & PEOPLE EATING PETS TO SURVIVE.TO FUTURE GENERATIONS: WE MEANT WELL
What Went Terribly Wrong
The funny thing is that this was not a denial of service attack. It wasn't a North Korean hacker trying to bring the west to its knees by disrupting this most vital of services. No, it was good old fashioned human error. I know it's hard to believe that with all of the super smart people at Google that it's possible that one (or maybe a few of them) screwed up, but apparently that's what happened.
The Official Google Gmail Blog post on the outage described the cause this way:
"... At about 12:30 pm Pacific a few of the request routers became overloaded and in effect told the rest of the system "stop sending us traffic, we're too slow!". This transferred the load onto the remaining request routers, causing a few more of them to also become overloaded, and within minutes nearly all of the request routers were overloaded."
What it Really Means
That approximately 80 percent of my tweets for 90 minutes involved this outage--so much traffic in fact, that ironically it could have taken down Twitter with it--tells you the extent to which people depend on this service. In spite of the fact that for most of us, Google's services are provided free of charge, we still go ballistic when it goes down. (Those who pay for it have a bigger gripe, but I'm willing to bet that's a small percentage of the people who were complaining.)
But regardless of whether you pay for Google services or not, the reaction (some might say overreaction) shows we have become dangerously dependent on them. We expect our services to be up all the time, and when they are not, we get REALLY upset. And when the collective web gets upset at the same time in the age of social computing, it can get ugly in a hurry.
What Should We Do?
We have to remember that computers are fallible. Networks are fallible and so is the genius of Google. Sure, Google has to put more contingencies in place, and test them regularly to make sure they work and this doesn't happen very often in the future. But we also have to realize that chances are, it will, and we have to learn to accept that. You have to ask yourself, how often do key applications go down inside the enterprise. How often do you experience issues on your internal networks?
These things happen when you're dealing with technology, so let's make a deal that the next time that Google goes down for a short time, maybe we don't have to discuss it ad nauseum on Twitter. Instead, perhaps we can all relax and enjoy the moment, maybe call a friend, go for a walk, spend time with family or send an email. Check that last one.
