Rockset For Developers

Rockset is a cloud-native, real-time analytics database. It lets users run low-latency, high-concurrency analytical queries on event streams and change data capture streams.
Developers love Rockset because:
  • It supports schemaless ingestion from a wide variety of data sources, including streams, databases, and data lakes
  • It’s fast. Rockset’s Converged Index(™) automatically indexes data three ways: column, row, and search, each of which is stored on RocksDB. Rockset executes queries against that data in 100s of milliseconds
  • It’s cloud native, fully managed, and easy to use, with no need for configuration or provisioning
  • It has robust SQL support, including search, aggregations, JOINs, and windowing, all on structured and semi-structured data….even heavily nested objects and arrays

Rockset: A 60 Second Overview:
Let’s say you’re building an anomaly detector to prevent credit card fraud and your analysis requires joining data from the following sources:
  • Real-time purchasing data, fed by an events stream
  • Customer data, stored in a NoSQL database
  • Historical transaction data, stored in a data lake
Preventing fraud saves your company money so performing analytics in real-time is business-critical.
Schemaless Ingest and Converged Index(™)
Create a collection, which is similar to a table but without a fixed schema, for each of your data sources. Rockset automatically ingests and indexes your data, including deeply nested DynamoDB objects, S3 fields with mixed types, and streaming Kinesis events. New data is available 1-2 seconds after it appears in your sources.
Fast, Familiar SQL
Now you need to calculate aggregate stats across sources. Write an SQL query which joins the data from your events streams, NoSQL database, and data lake. Get sub-second results on fresh data without learning another query language. Skip data transformations by leveraging SQL support for nested data.
REST Endpoints and Developer-First SDKs
Finally, you need to integrate these query results with an application. Create a Query Lambda, which is a dedicated REST endpoint that executes parameterized SQL, and invoke it with an SDK. Now your app can generate real-time alerts when anomalous events occur.
That’s It!
Because Rockset is cloud-native, it’s a very low-overhead product. Here’s a quick list of things you didn’t need to do:
  • Configuration
  • Cluster provisioning
  • Denormalization
  • Shard and index management

Go Deeper

Customers frequently compare Rockset with other real-time database products. We do the same, and we’re not the best choice for every use case. Check out some technical comparisons to see if Rockset is right for you: