Google Spanner
Google Spanner
Objective:
The primary goal of this paper is to inform us about Google's globally distributed database, Spanner.
Spanner is designed to be scalable and to offer high availability by replicating data over a worldwide
network of datacentres.
The database maintains data distribution and commutability amongst datacentres in order to
balance the load and respond to outages.
The spanner provides a variety of database capabilities such as a SQL-based query language, general
purpose transactions, and a semi-relational data format.
Servers: Spanners run on a large number of servers, which can be distributed globally or in separate
datacentres.
Network: A dependable network is essential for a spanner to connect with a server and a client.
Operating systems, cluster management software, database software, and application software are
all necessary software.
Cluster management Software: Spanner employs Google's cluster management software to keep
clock uncertainty to a minimum.
Database Software: Spanner is built on Google's architecture and includes capabilities such as
distributed transaction support and lock-free read-only.
Application Software: API applications that utilise spanner can interface with databases and use their
characteristics.
Major Contribution
The authors of the paper go into detail about the concept and implementation of Spanner.
The authors created Spanner, which uses the True Time Feature to create a distributed system with
stronger time semantics.
The author gives credit to Google personnel for their work on Spanner.
Pros and Cons
Pros:
Automatic shading, fault tolerance, reliable replication, and wide-area dissemination are also
advantages.
Cons