Real-Time Analytics on Oracle and MSSQL With Rockset

March 3, 2022

Register for
Index Conference

Hear talks on search and AI from engineers at Netflix, DoorDash, Uber and more.

Today Rockset is announcing an early access program for Oracle and Microsoft SQL Server integrations.

Oracle and Microsoft SQL Server (MSSQL) are both incredibly popular database products for transactional workloads at large enterprises. The amount of data companies generate, transform, store and query is growing exponentially. This data has material financial value when it’s both fresh and easy to access, however, customers commonly face scalability challenges running both transactional and analytical applications on the same database. This makes sense. Transactional databases must be write-optimized and analytical applications require low-latency reads.

Many architectures address this issue by sending data from transactional databases to a data warehouse — sometimes built on Oracle or Microsoft SQL Server databases. This tactic solves one issue, but at a high cost. Making data usable across systems often requires a slow, expensive, batch ETL/ELT process. rta-on-oracle-and-msql-image-1

Expensive, Slow Analytics

Moving data from an OLTP database to a data warehouse produces stale data — often hours or days old by the time it arrives. In general, there are two types of data consumers affected by poor data availability:

  1. Operational Analysts: These specialists analyze real-time data to find trends and insights to improve business observability. Delayed data means delayed responses and missed opportunities.
  2. Data Applications: Customers, both internal and external, interact directly with these applications. In this case, relevant, actionable information is not being served in real time.
  3. Enterprises have spent countless dollars reducing latency between transactional systems (powered by Oracle and Microsoft SQL Server) and data warehouses. Fast analytics on fresh data is better than slow analytics on stale data. Fresh beats stale every time. Fast beats slow in every space.

Real-Time Analytics on Oracle and Microsoft SQL Server Data With Rockset

Rockset is bringing fast analytics on fresh data to Oracle and MS SQL customers. Using Amazon Web Services’ (AWS) Database Migration Service (DMS), data from Oracle and Microsoft SQL Server is converted into a stream and sent to Rockset via AWS Kinesis.


Rockset tails the Kinesis stream, automatically indexes data and makes it available to analysts and data apps within a few seconds. Rockset is a fully-managed OLAP database that enables millisecond latency search, aggregations and joins on any data by automatically building a Converged Index™, which combines the power of columnar, row and inverted indexes. Rockset's Converged Index is the most efficient way to organize your data and enables queries on new data to be available almost instantly and to perform incredibly fast.

Rockset vs. Data Warehouses

  1. 10x Faster Queries: Rockset delivers sub-second query latency by indexing all your data for fast access. Data warehouses generally index a few columns and rely on time-consuming scans.
  2. Always Fresh Data: Preprocessing is required to improve query performance on data warehouses. This preprocessing incurs significant delays. Rockset ingests data in real-time, enabling analytics on fresh data.
  3. 50% Lower Compute Costs: Rockset enables compute-efficient analytics by minimizing scans and retrieving data exclusively from indexes. This eliminates the need for costly data transformations.


Join the Early Access Program

Sign up to join Rocket's early access program for Oracle/Microsoft SQL Server integrations. Our engineers will provide white-glove service during the setup process to ensure a hassle-free integration. We plan to work with a limited number of partners in the coming weeks before making the connector widely available for all our customers.

Sign up Now

Rockset is the real-time analytics database in the cloud for modern data teams. Get faster analytics on fresher data, at lower costs, by exploiting indexing over brute-force scanning.