According to Google, it’s the first database that can quickly store and retrieve information across a worldwide network of data centers while keeping that information “consistent” — meaning all users see the same collection of information at all times — and it’s been driving the company’s ad system and various other web services for years.
Spanner borrows techniques from some of the other massive software platforms Google built for its data centers, but at its heart is something completely new. Spanner plugs into a network of servers equipped with super-precise atomic clocks or GPS antennas akin to the one in your smartphone, using these time keepers to more accurately synchronize the distribution of data across such a vast network. That’s right, Google attaches GPS antennas and honest-to-goodness atomic clocks to its servers.
“It’s a big deal — and it’s really novel,” says Andy Gross, the principal architect of Basho, an outfit that builds an open source database called Riak that runs across thousands of servers — though not nearly as many as Spanner. “The conventional wisdom — at least among people with modest resources — is that time synchronization like that, on a global scale, that is accurate enough for such a big distributed database … just isn’t practical.”