Published July 21, 2022

Arrays, Tuples and Matrices, Oh My!: What’s in Ocient’s Version 20 

By Dylan Murphy, Director of Product Management at Ocient

Today, we’re excited to announce the version 20 release of the Ocient Hyperscale Data Warehouse. We see every new release as an opportunity to upgrade and expand our core product, and V20 is no exception.  

When planning the theme and features of this release, we identified an opportunity to help a growing number of customers endeavoring to analyze fast-growing datasets with a focus on log files and network metadata. In some cases, customers are using technologies that weren’t built for data analysis. In others, they’re working with legacy solutions. In both cases, they cannot ingest and analyze enough data fast enough to keep them compliant with industry regulations, monitor the health of their networks as closely as they’d like, and deliver critical insights to the business. 

Several of the features within v20 are critical to key use cases where we see Ocient delivering incredible performance gains while simultaneously cutting a customer’s total cost of ownership by 50% (and up to 80%) over three to five years. In some of these use cases, customers have evaluated cloud data warehouses and opted to stick with on-prem for increased security and control. But even if the customer is looking to deploy in the cloud, Ocient’s Compute Adjacent Storage Architecture (CASA) paired with the new suite of indexes and native complex data types launched with v20 have delivered even faster performance (up to 40x!) for these mission-critical workloads. 

What’s In v20?   

To give more insight into what’s included in our latest and greatest release, let’s unpack some of the features now available in v20 of the Ocient Hyperscale Data Warehouse: 

  • A new suite of indexes, including N-gram indexes, which enable Ocient to search text data significantly faster and more efficiently. Our N-gram indexes and new large string (large varchar) indexes have contributed to much of the performance increases now available in v20. To put our string indexes into context, Ocient can theoretically fit all text from the three Lord of the Rings books, PLUS the Hobbit, PLUS all seven Harry Potter books, PLUS all five Game of Thrones books into a single row stored natively as a string, then search through all of the texts in seconds. (Note to George R.R. Martin: there’s plenty of room for books 6 and 7). We’re excited to offer this incredible enhancement to customers for key use cases in telecommunications, government, and operational IT to find needle-in-the-haystack insights across a broad set of uses cases including call detail record (CDR) search, IP data record (IPDR) search, Internet connection record (ICR) search, and content delivery network (CDN) optimization.
  • Native complex data types, including arrays, tuples, and matrices, integrated as native components to bring even more performance improvements when ingesting and analyzing complex data types. Harnessing complex data types enables faster and more secure analysis of log and network data for multi-petabyte workloads.
  • ELT workloads. With the addition of CREATE TABLE AS (CTAS) and INSERT INTO SELECT, we’ve unlocked new opportunities to move data within Ocient. Taking the transformation step out of the loader nodes and putting it into the Ocient Hyperscale Data Warehouse makes it easier to distribute commonly used data, prototype new data designs, and deliver data to test environments and consumers.  

With v20, we’ve also released features that didn’t make today’s press release. In addition to our focus on analyzing log files and network metadata, our v20 release includes several features to enhance the experience of Ocient customers. 

  • SSO integration. V20 includes support for single sign-on (SSO) integrations, which allows customers to authenticate through their existing SSO environment. Ocient’s SSO implementation piggybacks on the database’s granular privilege and workload management functionalities. 
  • DDL improvements. Enhanced administration through data definition language (DDL) and improved system catalogs are designed to make it easier to integrate Ocient with customers’ environments.  As a result, customers have more ability to observe what’s happening within Ocient and even tie Ocient administration into their existing automation tools.   
  • Ocient Simulator. We developed a virtual testing environment for V20 that allows customers to test the features and functions of Ocient Hyperscale Data Warehouse without procuring the hardware required for a production system. Running in a Docker container, customers can interact with the Ocient Simulator directly on their laptops to explore new features and prototype solutions or run Ocient in continuous integration environments. Although the Ocient Simulator is not designed for performance or production use, it’s an attractive option for those looking to learn how to interface with Ocient.   

Whether you’re an existing customer looking forward to using our new indexes to ingest and analyze more data faster, or a prospective customer thinking about playing with the Ocient Simulator on your laptop, we can’t wait for you to check out Ocient Hyperscale Data Warehouse v20! 

Now, as we turn our attention to v21, we want to hear from you. What would you like to see in our next release? What challenges are you facing? Email us at info@ocient.com or provide your details via our Contact Us page.