Computerware Blog
Google’s Growth Spawns New Technology
At one point in Google's not-so-distant past, they were able to house all of their servers in one garage. True Story. Times have certainly changed for Google. In fact, Google has grown so rapidly that they saw the need to develop a new technology for the sole purpose of managing their growth. Google calls their new creation the Google Spanner.
The Google Spanner is a gigantic database that tracks all of Google's data. Google has a massive amount of data to keep track of. Consider just a few of Google's major ventures like YouTube, Gmail, Google Earth, Blogger, etc., each one of these services requires a data center to keep operations humming. Google is estimated to have more than a dozen data centers spread out all over the world, and one data center houses thousands of servers. Datacenterknowledge.com has estimated that, based on Google power consumption reports, Google has over 900,000 servers in operation.
The purpose of the Google Spanner is to manage the data on all 900,000 servers that are spread out across all of Google's warehouses around the world. This task is complicated further by the fact that the population of planet Earth is constantly accessing Google's data across 24 different time zones. To streamline the entire operation, Google Spanner not only keeps track of all of this data, but it also syncs everything with a singular time stamp.
The time stamp that Google Spanner creates for each piece of data is marked using True Time API, this time supersedes all of the other time zones. In order for the True Time API timestamp used by Google Spanner to be accurate, every one of Google's servers has to be fitted with both a GPS antenna and an atomic clock. If you are curious about the mechanics of the Google Spanner, then you can peruse a 14 page document published by Google. Here is an excerpt explaining how the timestamp works:
Data is stored in schematized semi-relational tables; data is versioned, and each version is automatically timestamped with its commit time; old versions of data are subject to configurable garbage-collection policies; and applications can read data at old timestamps. Spanner supports general-purpose transactions, and provides a SQL-based query language.
In layman's terms, this excerpt is explaining how the Google Spanner automatically balances the data demands across all of Google's servers so that retrieval time can be at a maximum.
The Google Spanner is not only designed to handle all the data that Google currently manages, but the Spanner also has the capability to manage all of Google's future data needs. The Google Spanner is designed to manage 10 million servers; this positions Google to grow more than 10 times their current size. Considering how Google is currently one of the world's largest companies, the thought of a Google that is 10x its current size is enough to give even George Orwell the willies.
Implementing the Google Spanner to your business might be overkill, seeing as your data needs do not likely exceed a room of servers; but like the Google Spanner, Computerware can streamline the use of your data. We have a process called virtualization that takes advantage of underutilized servers by condensing older server units into newer ones, this decreases the number of servers your business uses and saves you money on your electricity bill. Call Computerware at (703) 821-8200 and let us show you how you can get the most efficiency out of your IT infrastructure.
Comments