Google is reportedly considering commercializing Spanner, a highly scalable, globally distributed database technology that the company has been using internally for the past several years to deliver many of its core services.
Engineers from Google’s Cloud Platform Group are apparently working on figuring out how to turn the technology into a product that outside developers can use to build applications, The Information reported this week.
One of the biggest obstacles facing the company as it embarks on its mission is finding a way to decouple Spanner from Google’s proprietary hardware and network technology and building a version that will work just as well on other infrastructures.
According to The Information, Google’s efforts to commercialize Spanner are being driven by competitive forces. The company wants to be able to better compete with the likes of Amazon Web Services and Microsoft Azure in the cloud services space and is hoping that the highly-scalable, high-performance database technology will give it an edge over AWS and Azure, the report said quoting two unnamed sources who are apparently close to the company’s plans.
Google did not respond to a request seeking comment. The company has previously described Spanner as a SQL-like database that runs across multiple datacenters around the world and is capable of scaling up to literally millions of machines in hundreds of datacenters and trillions of database rows. According to Google, “it is the first system to distribute data at global scale and support externally-consistent distributed transactions.”
One of Spanner’s biggest features is its support for synchronous replication. In others words, changes made to Spanner data in one datacenter are automatically synchronized across all datacenters at the same time, so the data remains consistent regardless of where it is accessed.
“Applications can use Spanner for high availability, even in the face of wide-area natural disasters, by replicating their data within or even across continents,” according to Google.
One of the first Google applications to use the technology was F1, a massive relational database management system designed to support Google’s Adword’s ad management platform. Because F1 was based on Spanner technology, Google was able to use as many as five replicas of the RDBMS working in perfect concert with each other in datacenters around the country to ensure continuous availability of the platform.
Other applications that are thought to run on the technology or at least were touted as potential candidates for the technology include Gmail, Google Search and Picasa.
With Google not publicly talking about its reported plans to commercialize Spanner it is hard to say what kind of applications the company plans to run on the database.
One company that is working on releasing an open source version of the technology is Cockroach Labs, started by two engineers from Google. The company describes its CockroachDB, which is currently in beta, as being inspired by Spanner and Google’s F1.
According to Cockroach Labs, the database is a distributed SQL database that can scale horizontally and will be capable of surviving machine, disk, rack and data center failures with minimal application disruption and latency.
Cockroach Labs has said the database is especially well suited for applications that require a very high degree of survivability and horizontal scalability.