Sergey Nivens - Fotolia
SingleStore on June 23 released the latest update of its namesake product, a unified database platform that supports both transactional and analytical workloads.
The unified database vendor, based in San Francisco, was known as MemSQL until 2020 when it was rebranded as SingleStore. At the foundation of the SingleStore database is a capability that integrates columnstore and rowstore tables to help enable faster database queries.
That same model makes SingleStore effective for both online transaction processing (OLTP) and online analytical processing (OLAP) applications. The update from SingleStore provides new storage options for data, as well as performance improvements.
In a virtual launch event, David Hannibal, product and partnerships manager at IEX Cloud, explained how his firm's financial service cloud platform integrates SingleStore to help both OLAP and OLTP users.
IEX Cloud operates around the world and has more than 160,000 customers. Those customers build financial applications with financial data that IEX Cloud integrates with real-time and historic data. The challenge for IEX Cloud is being able to deliver data to users in less than 10 milliseconds. IEX Cloud uses SingleStore to help enable that fast data query delivery, Hannibal said.
"There are hundreds of content sets and hundreds of millions of data points that we have to maintain and support," he said. "We really did find a good, strong partner in SingleStore to help us deliver on that central tenet."
New updates improve unified database platform features
In an interview, SingleStore Chief Product Officer Jordan Tigani noted that a key update in the new release is the expanded storage capability.
Tigani explained that SingleStore has always had a two-tier storage hierarchy. The first tier is memory, where the original MemSQL name originates. The second tier is local disk, or persistent disk, which is, in the cloud space, often solid-state disk. With the update, SingleStore is adding a third tier with cloud object storage, which includes services such as Amazon Simple Storage Service.
David HannibalProduct and partnerships manager, IEX Cloud
A key challenge with cloud object storage for databases is that it's slower than an in-memory approach. SingleStore data storage tiering first writes updates to the fastest in-memory tier and then copies out to the cloud object storage tier to help improve overall performance.
Multicolumn keys boost SingleStore query performance
Another feature that has landed in the update is multicolumn key support.
Tigani explained a typical database includes a primary key that defines where and how data is stored. The primary key is how fast data lookups in a database are enabled, as it provides direction as to where data is located.
Previously with SingleStore, keys were limited to a single database column. With the new support for multicolumn keys, more complex queries can be executed quicker, as the keys can be stored in multiple columns.
Looking forward, Tigani said SingleStore will continue to build out its technology as a high-performance unified database platform that also has multimodel capabilities such as document database and geospatial capabilities.