Google Cloud Spanner is a fully managed, mission-critical relational database service that offers transactional consistency at a global scale, automatic synchronous replication, and support for SQL dialects. It’s designed to combine the benefits of relational database structure with non-relational horizontal scale.
Architecture
Spanner uses a columnar storage schema, which allows for fast data access and navigation. The data is stored in tables that are partitioned by geography and time. Spanner also uses an indexing strategy that allows for fast searches.
A Spanner deployment consists of a number of nodes, which are units of compute capacity. These nodes serve read and write/commit transaction requests, but they don’t store the data. Each node is replicated across three zones in a region.
The actual data is stored in Google’s underlying distributed file system, known as Colossus. Colossus takes care of data replication, recovery, and distributed access.
Consistency and Replication
Spanner provides strong consistency at a global scale, which is a significant achievement in the world of distributed databases. It uses the Paxos consensus algorithm to ensure that all copies of the data agree with each other.
Spanner also uses synchronous replication for writes. This means that when a write occurs, it is immediately replicated to all copies of the data. This ensures that the data is always up-to-date, regardless of where it is accessed from.
TrueTime API
One of the unique features of Spanner is its TrueTime API, which allows it to achieve strong consistency at a global scale. TrueTime provides a globally synchronized clock, which is used to assign timestamps to transactions. This allows Spanner to order transactions in a way that is consistent across all replicas.
SQL Support
Spanner supports a dialect of SQL for querying data. This makes it easy for developers to work with, as SQL is a widely known and used language in the world of databases.
In conclusion, Google Cloud Spanner is a powerful tool for managing large-scale, globally distributed data with strong consistency and high availability. Its unique features like the TrueTime API and strong SQL support make it a compelling choice for many organizations.