Database startup ScyllaDB Inc. is using its two-day Scylla Summit starting today to announce the release of Scylla Open Source 3.0.

The new product represents a major upgrade to ScyllaDB’s open-source NoSQL database, with a host of new features designed to achieve parity with the Apache Cassandra database with which it’s trying to compete.

Scylla’s database is actually derived from Apache Cassandra, which was first developed by Facebook Inc. as a distributed, wide-column store, NoSQL database management system for handling large amounts of data across many commodity servers. What ScyllaDB did was rewrite the database in a different programming language that allowed it to make several changes under the hood to boost its performance. The company claims that Scylla can now run 10 times faster than Cassandra with lower latency thrown in for good measure.

ScyllaDB said its database is being used in production for numerous specialized tasks by big organizations such as IBM Corp. and the European Organization for Nuclear Research or CERN. Some use the platform to process time series data, which refers to a series of data points that are indexed, listed or graphed in time order.

Meanwhile, it’s being used as a graph store by others. Graph stores allow users to pinpoint relationships between different data points, which is useful for data analytics and security diagnostics.

The main point of today’s upgrade is to add new capabilities to Scylla that were previously not ported over from Cassandra. There are also some entirely new features, such as support for concurrent online transaction processing or OLTP and online analytical processing or OLAP, enabling these to be done simultaneously for the first time.

This latter feature is important because the two analytics methods access data in different ways. OLTP involves analyzing small and varied transactions including mixed writes, updates and reads, with a high sensitivity to latency. OLAP is more about broader scans across multiple datasets.

Some of the other new features include “materialized views,” which is an experimental feature since dropped from Cassandra that enables automated server-side table denormalization, that last an approach to speeding up read-oriented data retrieval performance. Scylla also gains support for “secondary indexes,” which enable data to be queried through nonprimary key columns, and support for the Apache Cassandra 3.x compatible storage format, which the company said boosts both performance and storage volume by up to three times.

“We finally reached parity with Apache Cassandra, and developed a groundbreaking OLTP+OLAP service level agreement guarantee that puts us on a path toward pure multitenancy,” Dor Laor, ScyllaDB’s co-founder and chief executive officer, said in a statement.

The company said Scylla Open Source 3.0 will be made available later this month.

Image: Sasint/pixabay