Quickly visualizing and analyzing data is key to creating a data-driven culture. Speed improves productivity and creativity and helps you focus on business problems. Oracle Analytics Cloud includes a new in-memory technology that provides in-memory caching, where data is optimized and compressed for much faster performance for large datasets. With this technology, you see 2-5 times improved query performance with some queries improving by 20 times or more!
In Oracle Analytics Cloud, the new in-memory engine provides Automatic Caching, which is built on the Oracle Database in-memory compute technology. Because it's built on Exadata hardware and the advanced Oracle Database query engine, Automatic Caching for datasets offers query performance improvements without any changes to your datasets. The in-memory technology is especially helpful for larger datasets or datasets with complex calculations.
While in-memory technology is becoming commonplace in the analytics industry, Oracle Analytics Cloud’s approach leverages all of the power of the Oracle in-memory database compute technology. This new acceleration cache supports complex data types and advanced features such as geometry data types, complex joins, and the full range of Oracle Database capabilities. The performance of queries in Oracle Analytics Cloud improves with the number of users running queries as the system learns from and optimizes based on real-world usage.
The ability to quickly answer questions is at the heart of self-service analytics. Speed is essential to enable users to focus on the business problem and not the tool. Business users can ask new questions and quickly get answers without worrying about optimization and data management. The ability to do fast filtering, including capabilities such as partial string matches on text, enables business users to quickly and easily answer questions, which would have required data preparation or new datasets in the past.
Oracle Analytics Cloud In-Memory Engine
The new caching engine is available in Oracle Analytics Cloud in Oracle Cloud Infrastructure (OCI). When users reload datasets in the latest release, the Automatic Caching engine leverages this new in-memory technology. The new engine has a hybrid columnar and row store to enable the fastest and richest query semantics. Oracle Analytics Cloud’s in-memory technology uses the latest, deep technology such as SIMD vector instructions and Exadata cell-off loading for optimal performance. In earlier Oracle Analytics Cloud releases, users in logistics saw typical queries running in a range from 0.9 second up to 180 seconds. After the rollout for the same query patterns, the queries ranged from 0.6 second to 3.2 seconds. With the new engine, 83% of queries return in less than 2 seconds with a 30% increase in sub-second queries.
Some of the best performance improvements are for more complex scenarios. Datasets with multiple users using shared workbooks see query performance improving as more queries are run. Users often see the most benefit when employing complex filters and calculations.
For example, the following figure shows a comparison of a cash-on-hand forecast by month for the next 12 months showing a 20 times query performance improvement.
These performance benefits come with the same security and ease-of-use that you expect with an Oracle Analytics Cloud dataset. For example, you can combine Automatic Caching with a blend of datasets from different data sources, and the mashup has the same blazing performance. Whether your data comes from an Oracle Application such as Oracle Fusion ERP or Taleo, from a source such as Google Drive or Dropbox, or an on-premises database using the remote data gateway, the new caching engine can bring in-memory technology and performance to those datasets. You can also schedule reloads of the datasets to ensure data is up-to-date for your business needs. Beyond self-service data visualization workbooks, you can use these datasets with capabilities such as dashboards and OA Publisher reports scheduled with Delivers, utilizing the enterprise capabilities of Oracle Analytics Cloud. All of this is built on the world-class cloud infrastructure and security of the Gen2 Oracle Cloud Infrastructure (OCI).
This new engine is the foundation of a range of new features, which will be released in Oracle Analytics Cloud over the next year, including advanced capabilities such as spatial analysis and advanced analytic functions such as market basket analysis. Sharing this core technology with the Oracle Database team also means that Oracle Analytics Cloud will keep innovating and adding deep capabilities to the engine.
The best part of this new capability is that customers already have it. Starting with Release 6.1, all customers using Oracle Analytics Cloud on OCI Gen2 have been automatically upgraded. Simply continue to use datasets with Automatic Caching, and when the data is reloaded, it will use in-memory technology.
Source: oracle.com
Posts
0 comments:
Post a Comment