ClustrixDB

Extreme scale, no compromise

ClustrixDB is a relational database engineered from the ground up to power web, mobile and Internet of Things (IoT) applications with the most extreme scalability requirements, and to do it without compromising on the critical features databases need to provide mission-critical applications with standard access to reliable data: transactions and SQL.

Use cases

  • Ad Tech
    Ad Tech

    ClustrixDB provides digital ad platforms with the scalability to store billions of tracking events and deliver millions of real-time ads.

  • E-commerce
    E-commerce

    ClustrixDB helps E-commerce leaders scale to provide a great customer experience to millions of shoppers through billions of interactions.

  • Gaming
    Gaming

    ClustrixDB helps online and mobile gaming platforms scale to handle millions of players expecting real time feedback – and no “lag”.

  • SAAS
    SAAS

    ClustrixDB makes it easy for SAAS providers to scale out fast, frequently and far while growing from a single customer to many millions.

  • Social
    Social

    ClustrixDB equips web and mobile social apps to connect millions of people globally with real time messaging and status updates.

  • ClustrixDB is a holy grail of sorts. Its ability to scale out was something we couldn’t achieve with MySQL…

    David Smith, CTO, Viverae

  • With ClustrixDB, we stopped running into scaling issues, seamlessly growing our cluster from 3 nodes to 21 nodes.

    Nicolas Van Eenaeme, CEO, Massive Media

Architecture

  • Distributed Data
    Distributed Data

    ClustrixDB distributes table “slices” across the cluster with consistent hashing, rebalancing the data when nodes are added/removed.

  • Distributed indexes
    Distributed indexes

    ClustrixDB maintains distributed indexes, rather than local indexes, to improve performance at scale by avoiding scatter/gather overhead.

  • Distributed queries
    Distributed queries

    ClustrixDB executes queries with Sierra, a query optimizer based on the Cascades Framework sitting on a distributed execution engine.

Scalable relational

  • Standard SQL
    Standard SQL

    ClustrixDB implements standard ANSI SQL, including joins, aggregates and JSON functions, and is compatible with MariaDB connectors.

  • Strong consistency
    Strong consistency

    ClustrixDB uses two-phase locking, MVCC (snapshot isolation) and consensus (Paxos) to guarantee strong consistency at any scale.

  • High availability
    High availability

    ClustrixDB distributes tables “slices” and automatically replicates them to ensure data remains available if one or more nodes fail.

Unlimited scaling on AWS

ClustrixDB provides unlimited read, write and storage scaling on AWS – something Amazon Aurora cannot do. It is constrained by a shared-storage architecture, and while Amazon Aurora can scale out reads with up to 15 replicas, it is limited to scaling up writes with a single master. ClustrixDB is compatible with MySQL clients, but unlike Amazon Aurora, it’s distributed, scales out writes and does not result in separate IO and storage fees.

Helpful resources