Database. Queue. Cache with secondary indexes.
Tarantool is an open-source database that can store everything in RAM. You can use it as a cache with on-disk persistence. It is able to process up to 1 million requests per second, supports SQL and secondary index searching.
In Tarantool, you can execute code alongside data. Lua programming language helps you implement any business logic and speeds up your solution. Get rid of the outdated database records. Synchronize with other data sources. Implement HTTP-service.
To implement business logic that requires atomicity.
To balance the server load and duplicate the data to keep it safe.
To snapshot your data on disk automatically to restore it in case of failure.
To apply ready-to-use queues, connectors to other databases, and other extensions.
To use ANSI SQL with the JOIN and LIKE clauses support.
To store your hot data in RAM and cold data on disk.
If you use sharding and replication, you need a cluster. Using cluster provides fault tolerance, makes it easier to configure topology, and brings a friendly error reporting via UI. Organize your Tarantool instances into a cluster and develop distributed applications with the help of Tarantool Cartridge.
We needed a fast storage with 50+ million records capacity and secondary indexes support. When using PostgreSQL, we faced some issues. Sure, there must be a way to implement the case with PostgreSQL, but I did it faster by using Tarantool.
It's nice to see we have an excellent domestic product that is besides other things suitable for developing highload systems. A product that is fast, scalable, and reliable.
Tarantool's high performance and accessibility at any level as well as the wide range of capabilities of the embedded language and application server make this platform a truly multi-purpose and powerful tool for solving OLTP tasks of any kind.
Two DB engines, memtx (in-memory) and vinyl (on-disk), combined with the capability of developing business logic proved to be a real godsend to us. We've been using Tarantool since 2018 with terabytes of data and dozens of instances in the cluster. Works 100%.
High performance. High availability. Flexibility when using in web applications (data storages, queues, proxy servers, RPS limiters, sharding, billing, caches, integrations).
The best customer support I've ever dealt with.
of developing data storing technologies
users around the world
of supporting high performance services