Actian helps Evernote deliver on promise to help customers “Remember Everything”

REDWOOD CITY, USA: With more than 80 million users globally and growing, Evernote knows a thing or two about architecting for sustainable, rapid data proliferation.

When the time came to do a complete redesign of its solution to handle data that was bursting at the seams - more than 40 billion rows of data in the primary table - Evernote implemented its big data analytics using the Actian columnar MPP database which, as Evernote CTO Dave Engberg notes in a blog, "excels at superfast ad-hoc queries."

"Evernote acts as a second brain to many of our users. They expect to be able to recall details from their accounts instantly anytime, anywhere and we depend on super-fast data retrieval to deliver that experience," says Engberg. "As we began our next-generation architecture, we focused on another value angle - being able to conduct meaningful analytics on accumulated data, including large-scale aggregate processing and calculation across many independent variables."

True to the spirit of innovation that pervades Evernote, Engberg's team completely reinvented their approach to analytics with a two-tier environment leveraging both Hadoop and the Actian Analytics Platform.

The Hadoop cluster, accessed via Hive, absorbs the rapidly accumulating log files, accommodating affordable, massive data storage and parallel batch processing jobs. The data, however, is cloned into the Actian analytics database for more complex, multifaceted analytics. There, as Engberg notes: "We're able to answer many types of questions much faster than we could with a brute-force crawl in Hive. We get a great performance boost with the Actian analytics database."