Normally, businesses have to pick between a traditional database or one which has an SQL database. Whichever they choose, they’ll have to sacrifice one thing to get the other. With a traditional database, there’s better consistency but difficulty with scaling. With an SQL database, it’s the exact opposite — easier to scale but less inconsistency.
Cloud Spanner is designed to integrate both database capabilities so you won’t have to make a choice on which you prioritize more. It will be offered as a fully managed service, with Google (NASDAQ:GOOGL) managing everything including hardware, software and content replication, regardless of whether the database spans a single data center or several data centers. In other words, as Google takes care of all the hard stuff, all you have to do is rent out Spanner and it’s just like running apps and services using the very same technology that Google itself is using.
Cloud Spanner took off in 2007 when a group of Google engineers decided it was time to end the dilemma of having to choose between an SQL and a nonSQL database. This led to the creation of Spanner which was described in a paper published in 2012 as a globe-spanning database that stores information from machines in data centers across several continents.
Its size alone is overwhelming enough. What makes it even more impressive is how it manages changes in data. Specifically, changing data in one part of the database will not contradict other changes done on some other part of the database. Plus, it can easily replicate data across several data centers in different regions of the world, and readily retrieve these copies in case of data center outages. This extraordinary capability that only Mountain View has, was made possible through Google’s timekeeping technology which they aptly called TrueTime.
Before TrueTime, one of the biggest challenges with databases was consistency — communicating across the globe didn’t just take too long, time differed from machine to machine too. This made it impossible to ensure that transactions in different parts of the world aligned with each other. As a result, seamless data replication cross regions became even harder.
But by installing atomic clocks and GPS receivers within their data centers, Google was able to synchronize those different times on different machines so that the entire database could have one common time and in the process enable data to flow in or out in the proper sequence. And by incorporating TrueTime in Spanner, Google has built a database that’s on an entirely different level compared with other existing databases, providing this way some added leverage in its battle with software giant Microsoft (NASDAQ:MSFT) and Amazon.com (NASDAQ:AMZN) for supremacy in the cloud computing market.
Pricing for Cloud Spanner, which is now open to everyone and accessible via the Google Cloud Platform (GCP), is dependent on the amount of nodes, storage and bandwidth used. It starts at $0.90 per node per hour, while storage costs $0.30 cents per gigabyte per month. Ingress (incoming) traffic is free while egress (outgoing) traffic is charged according to current commercial rates.
Disclaimer: This page contains affiliate links. If you choose to make a purchase after clicking a link, we may receive a commission at no additional cost to you. Thank you for your support!
Anxious to start hearing about experience moving large application from Oracle and DB2 to Spanner. This looks almost too good to be true.
This is only news because it’s Google. There are other databases such as VoltDB (created by Turing award winner Dr. Stonebraker) that is also a consistent and scalable database.