Introducing deepSQL

Applications and their performance drive your business.  Unfortunately, even modern databases are disconnected from application needs, falling short on performance and scale while requiring constant manual tuning...until now. 

deepSQL is an application-aware relational database designed for Big Data and the Cloud. deepSQL combines our CASSI based Big Data engine with the popular MySQL interface to reimagine the relational database. It concurrently runs low-latency data streaming, transactions and analytics.  Through a unique approach called CASSI, machine learning algorithms make intelligent decisions about how, when and where deepSQL conducts operations based on unique application workload characteristics combined with the capabilities of the hardware. The result is a virtually lockless database that linearly scales based on CPU and memory. Concurrent workloads accelerate to in-memory performance levels with full ACID compliance. 

deepSQL makes your database application-aware by predictively meeting your unique application needs enabling unprecedented performance for Big Data in the Cloud.

Multi-Dimensional Architecture
Using independent in-memory and on-disk data structures, machine learning algorithms and intelligent resource scheduling, deepSQL minimizes data overhead and maximizes hardware resource utilization. Techniques such as dynamic virtual data representation enable deepSQL to predictively organize and optimize data on-the-fly in memory, enabling optimization for each workloads unique needs in the same database.

On-Demand Analytics on Operational Data
deepSQL employs hyper indexing and dynamic summarization to embed critical analytical meta-data (statistics, cardinality, counts, etc.) into memory, allowing real-time analytics to run on live operational data as it streams into the system. This empowers applications to make mission-critical decisions in milliseconds rather than minutes, even in complex scenarios that require multiple joins, unions, etc.

Performance at Cloud Scale
deepSQL scales-up database tables into the hundreds of billions of rows, expands scale-out capabilities and accelerates replication on physical, virtual, cloud, or converged infrastructures. Based on real-time OS principles, deepSQL performs highly parallelized operations, removing IOPS as a barrier to scaling up. Scalability becomes linear based on CPU and memory, while system performance is maintained during extreme user and workload concurrency, without manual human intervention.

deepSQL: The Application-Aware Database

deepSQL: The Application-Aware Database

Easy to Deploy, Familiar Interface
deepSQL uses the well-known MySQL API with JSON support and is fully MySQL compliant. Developers, DBA’s and data architects get the best of MySQL at NoSQL scale and continue to leverage the vast MySQL ecosystem out of the box. Existing MySQL applications work seamlessly and new applications can be developed quickly based on rich MySQL capabilities that scale.

Cloud Automation Platform Integration
deepSQL enables Database as a Service (DBaaS) through integration with popular cloud automation platforms including BOSH, Cloud Foundry, Chef, vCloud Director, Vagrant, Compose, OpenStack and others. Dynamic Resource Awareness instantly utilizes any changes to the assigned resources (CPU, memory, etc.) with zero service disruption, empowering elastic scale without human intervention on supported virtualization platforms (VMware ESXi & Photon, KVM, Linux Containers, Docker (april)).

deepSQL Performance

  • 80%
    reduction in IOPS
  • 5x
    or faster ingest with heavy indexing
  • 32%
    greater tenant density
  • 50%
    or more compression, no performance impact
  • 4x
    or faster query performance than INNODB
  • 50%
    lower machine costs
  • 4x
    or faster page loads
  • 10x
    or faster transactions