Amazon Aurora, a New Distributed Relational Database from AWS
AWS launched a MySQL compatible databases called Amazon Aurora at AWS re:Invent this morning. The database is designed to be fault tolerant, scalable and an alternative to to high-end SAN environments.
It is compatible with MySQL and can scale to run six million inserts per minute. It is available through the AWS RDS service.
The relational database will add storage in 10 gigabyte increments on an as-needed basis. It can scale up to 64 terabytes. Storage scales in a linear fashion as more data is stored.
According to the AWS blog, storage is replicated across three AWS availability zones with two copies of the data in each availability zone. Aurora also has concurrency and self-healing built in and is distributed across SSD-powered storage.
“Amazon Aurora can tolerate the loss of two copies of the data while it is handling writes and three copies of the data while it is handling reads.”
According to AWS, with this method, data is backed to Amazon Simple Storage Service (S3), which is done in parallel without imposing any load on the database instance. Instant backups are automatic.
Aurora is what we see often described as a converged system. It combines the “storage, network, compute, system software, and database software.”
The news from AWS today is another example of the shift to new stack architectures that can take advantage of distributed infrastructures that their predecessors just can’t match.