The Oracle Cloud@Customer team relies on Autonomous Database to deliver analytics through Oracle APEX, Oracle Analytics Cloud, and web applications.
“We are experts in data science and analytics, but without deep database administration knowledge. Because Autonomous Data Warehouse manages the system for us, we were still able to build a successful data warehouse with over 60 data sources and support for many different users.”
Oracle Cloud@Customer supports customers who want to move to the cloud, but don’t want to move out of their data center. Keeping that business running smoothly requires extensive analytics on data gathered from device telemetry and more than 60 more different sources.
The original implementation required team members to manage the data warehouse themselves. As experts in data science and analytics, this was a task they didn’t want to do, so they looked for other options.
Why Oracle chose Autonomous Database
The data science and analytics team was among the first adopters of Autonomous Database when it became available. By relying on the database to do the hard work of managing backups, updates, tuning, and more, the team members were able to focus on building the right analytics and making that available to their key stakeholders.
Today the Cloud@Customer data warehouse is just over 47 TB in size and runs on 20 OCPUs. It ingests data from more than 60 different sources, and supports the business needs of the executive management team and operations (including patch management, support, and more). It offers a range of ways to access the data, including Oracle Analytics Cloud, custom no-code applications built with Oracle APEX, web applications, and Oracle Data Visualization.
The data warehouse supports some critical capabilities, so it is protected with Autonomous Data Guard, which was a 5-minute exercise to set up for intra-region failover.
Autonomous Database performs automated database backups, helping the data science and analytics team to easily recover tables and data as needed. As usage grew, the team did need to allocate more resources. Growing the database from 10 to 20 OCPUs took just a few clicks with no preparation, no downtime, and no database expertise required.
With autoscaling enabled, when usage spikes temporarily the database adds more resources to preserve performance. In typical usage, autoscaling kicks in about once a week, adding up to 10 additional OCPUs. Without autoscaling, guaranteeing the same performance all the time would require permanently provisioning 30 OCPUs (to be ready for that short period of peak usage) so autoscaling provides a valuable cost savings as well as a better user experience.
The team can also use data science tools to create models using the data in Autonomous Database. As data scientists and analysts, the team wanted to spend time creating new functionality, not administering a database. Autonomous Database has enabled exactly that.