Citus gives you the Postgres you love, plus the superpower of distributed tables. 100% open source. Now with schema-based and row-based sharding—plus Postgres 16 support!
Scale Postgres by distributing data & queries. You can start with a single Citus node, then add nodes & rebalance shards when you need to grow.
Parallelized Performance
Speed up queries by 20x to 300x (or more) through parallelism, keeping more data in memory, higher I/O bandwidth, and columnar compression.
Power of Postgres
Citus is an extension (not a fork) to the latest Postgres versions, so you can use your familiar SQL toolset & leverage your Postgres expertise.
Simplified Architecture
Reduce your infrastructure headaches by using a single database for both your transactional and analytical workloads.
Open Source
Download and use Citus open source for free. You can manage Citus yourself, embrace open source, and help us improve Citus via GitHub.
Managed Database Service
Focus on your application & forget about your database. Run your app on Citus in the cloud with Azure Cosmos DB for PostgreSQL.
Citus = Postgres At Any Scale
The Citus database gives you the superpower of distributed tables. Because Citus is an open source extension to Postgres, you can leverage the Postgres features, tooling, and ecosystem you love. And in Citus 12, thanks to schema-based sharding you can now onboard existing apps with minimal changes and support entirely new workloads like microservices. With Citus, you can scale from a single node to a distributed cluster, giving you all the greatness of Postgres—at any scale.
SaaS apps often have a natural dimension on which to distribute data across nodes—dimensions such as tenant, customer, or account_id. Which means SaaS apps have a data model that is a good fit for a distributed database like Citus: just shard by tenant_id—and for cases with no natural distribution key, you can use schema-based sharding.
Customer-facing real-time analytics dashboards need to deliver sub-second query responses to 1000s of concurrent users, while simultaneously ingesting fresh data and enabling users to query the fresh data in real time, too.
By scaling out Postgres across multiple nodes, Citus gives your analytics dashboards the compute, memory, and performance they need to process billions of events in real time.
Whether your app works with financial data, website analytics, IOT data, ad tech, or any other type of monitoring data—time series workloads often need to grow beyond the memory, compute, and disk of a single node.
Citus gives you parallelism, a distributed architecture, and columnar compression to deal with large data volumes—and can be combined with PostgreSQL’s time partitioning to optimize your timeseries queries.
Features for Time series workloads
Sub-second query responses for billions of events
Citus columnar for compression of older data
Integrated with native Postgres partitioning (& pg_partman)
Citus supports schema-based sharding, which allows distributing regular database schemas across many machines. This sharding methodology fits nicely with typical microservices architecture, where each microservice can have its own schema in a shared distributed database.
Schema-based sharding is an easier model to adopt, create a new schema and just set the search_path in your service and you’re ready to go.
Advantages of using Citus for microservices:
Ingest strategic business data from microservices into common distributed tables for analytics
Efficiently use hardware by balancing services on multiple machines
See how Citus gives this application ~20X faster transactions and 300X – 150,000X faster analytics queries.
A side-by-side comparison of Citus vs. single-node Postgres, comparing the performance of transactions, analytical queries, and analytical queries with rollups.
Customers ask how many weeks they have to wait for results, but with Citus it takes us roughly three minutes. The difference is so striking compared to our competitors, customers usually struggle to understand how we can do it.
We expected a database migration to be a huge code change, requiring a lot of rewrites. But Citus has a Django multi-tenant library, which minimized the refactoring to migrate our application. Except for rewriting a few lines of our app, we ended up using pretty much the same code.
Citus handles our complex ingest and delete operations seamlessly—and that’s with a workflow that creates five to ten billion new rows of data per day. The way in which Citus shards and distributes data across multiple nodes has given us the parallelization we need.
Sylvain Friquet, Former Software Engineer, Algolia
We provisioned a new database using the Citus console, ran our migrations, copied over our data from Heroku, and we were up and running on Citus in just a few minutes. And the technical benefits were instant. We saw queries that were once taking 15-30+ seconds reduced to less than 500 ms.
