Amazon Web Services launches Amazon MemoryDB for Redis, a real-time database implementation that is partly derived from the open-source Redis platform but with more durability and persistence.
This new product is aimed at customers who need a full database service, rather than a cache for Redis, based on the assumption that the use cases will be different.
Versatility has long been a feature of Redis’ success. It is most commonly used as an in-memory cache that works with a transactional database such as SQL Server, MySQL or Oracle.
The need for a Redis database has led AWS to introduce MemoryDB – although the capabilities of MemoryDB and Redis Enterprise differ greatly in this respect.
Redis Enterprise offers tiered storage to Flash, a feature not included in the original version. However, a major difference from MemoryDB is it approach to data persistence. It uses a multi-AZ transaction log to keep data permanent and allow replacement and recovery without data loss. The log does not run on MemoryDB database nodes.
Amazon MemoryDB supports all central data APIs of open source Redis, including strings, lists, sets, sorted sets, hashes, streams, geospatial data, bitmaps and hyperloglogs. It supports up to 128 Tbytes of memory per cluster, with a replica for each shard.
By offering a service that AWS calls Redis compatible and that supports the open-source Redis APIs, there are similarities with AWS approaches to Amazon Aurora for MySQL and PostgreSQL and Amazon KeySpaces for Apache Cassandra. In addition, they support the same APIs as the popular open-source data store with improvements under the hood to provide a fully managed experience.
Unlike most AWS database innovations, which usually start with a preview, MemoryDB is now available to everyone.
For more information, read the original story in ZDNet.