Tarantool for a scalable system

Tarantool for a scalable system

It's easier and cheaper in the long term if service architecture initially allows for scalability. Tarantool can be easily deployed on a single node, and as the load increases, the platform adds more nodes automatically to distribute the data evenly among them. Install Tarantool and manage loads of any size.

Tarantool for a scalable system

When to scale

Today, it's best practice to provide for scalability at the very outset of development. For this reason, you might want to rely on additional instruments from the very beginning of your project. Vertical scaling is strictly finite, and horizontal scaling is complicated and expensive if your system operates under load—there are no out-of-the-box solutions for that.

Thus, it's most viable to make room for horizontal scaling at the very outset of the project. In fact, this doesn't cost you anything—launch nodes on a single machine and scale them across multiple when it's necessary. You can easily do all that using our in-memory platform. It has everything you would expect from a relational DBMS—schemas, SQL requests, and ACID transactions, all tested and ready for global scaling.

Developing a project from scratch

Running out of project storage

Cache overflow risks

Get a scaling solution for your running service

Avoiding risks

You can use Tarantool as early as when creating an MVP. In this way, you'll avoid many problems associated with quickly growing relational databases, such as decreased throughput and lags. RDBMSs don't provide any ready-to-use instruments for horizontal scaling. Of course, your cache will be able to handle the increasing load for a while. But when the cache becomes too large, you may face problems associated with invalidation, cold cache, and consistency—which usually leads to increased costs.

Tarantool helps avoid these problems. It can act as a regular relational DBMS in your prototype. However, Tarantool is efficient in data-intensive systems and supports horizontal scaling based on sharding. It is universally applicable, so you won't need to repeat the same steps with another instrument later.

Database and application platform, all in one

scrollable diagram
risks diagram

Why use Tarantool?

High speed

Every node can process up to 1 million requests per second

Horizontal scaling

Built-in sharding and clusterization

Resource optimization

Create nodes on one or multiple machines and scale to 1,000 clusters in 40 minutes

Reliable storage

Synchronous replication and transaction logging

Tell us all about your system, and we will help you adopt Tarantool quickly and with minimal costs

a consultation

a demo

Thank you for your request

Tarantool experts will contact you shortly