Oracle Autonomous Database features free DBAs from routine tasks
Oracle Autonomous Database can automate routine administrative and operational tasks for DBAs and improve productivity, but migrating to the cloud has its share of issues.
Oracle Autonomous Database features and related cloud services can help automate administrative and operational tasks for IT teams, freeing up database administrators and systems managers from routine tasks so they can focus on more complex projects -- but potentially also enabling organizations to reduce their ranks.
The autonomous database service, which became available in 2018, combines Oracle's flagship database software with added automation features driven partly by machine learning algorithms. As a result, DBAs don't have to do as much -- or any -- hands-on work to deploy, configure and manage databases.
"The Autonomous Database eliminates the non-value-added tasks of the DBA, such as patching, tuning and query optimization," said Tony Baer, principal at database and analytics consultancy dbInsight. Although Baer said he sees many early adopters of Oracle Autonomous Database reducing or redeploying DBAs, they "will not be eliminated, as you will still need them for the strategic tasks of planning databases, laying out schema and policy implementation. But you will likely not need as many of them."
Oracle's Autonomous Linux technology could also help organizations update servers in response to newly discovered security vulnerabilities, said Robert Shimp, group vice president of infrastructure technologies at Oracle. Introduced in September 2019, Autonomous Linux can apply security patches in online fashion without bringing down applications or servers.
In addition, autonomous capabilities help eliminate human error or delay from the cybersecurity lifecycle. Automation at the OS level can dramatically improve the security and uptime of these systems, at scale and without human intervention, by automatically addressing threats promptly when a patch is available, Shimp said.
Tony Baer
Autonomous Linux complements the two versions of Oracle's Autonomous Database that are available for online transaction processing and data warehousing. The database itself encrypts all data and provides automatic security updates with no downtime, according to Oracle. Using tools like Oracle Real Application Clusters and cross-region Oracle Active Data Guard, the database supports self-recovering capabilities to automatically detect and apply corrective actions in response to problems, with Oracle guaranteeing 99.995% uptime overall.
Planning for migration
[Y]ou will still need [DBAs] for the strategic tasks of planning databases, laying out schema and policy implementation.
Tony BaerPrincipal, dbInsight
Enterprises need to consider many potential issues in moving to Oracle's Autonomous Cloud services, including reigning in out-of-control operational and hardware costs as well as overextended workloads for DBAs.
Currently, Autonomous Database is available only in the public cloud. The database runs on the Oracle Exadata system, and Oracle previewed a version for its Gen 2 Exadata Cloud at Customer platform at Oracle OpenWorld 2019. That will enable customers to run Autonomous Database in their own data centers with the vendor managing it for them, but Oracle hasn't publicly announced when the on-premises support will become available.
Moving to Oracle's Autonomous Cloud services requires a sound database design. If the schema isn't optimized for the query or transaction workload, the savings from running on Oracle's Autonomous Database will be limited. "You might not need as many DBAs, but an autonomous database is not a replacement for a smart DBA," Baer said.
There are two basic strategies for moving a database to the cloud. The simplest strategy is lift-and-shift, in which the workload is migrated to the cloud unchanged. Another strategy, Baer said, is lift-and-transform, in which the cloud migration allows a company to rethink and optimize workloads and database designs.
The ideal migration should proceed in a phased manner, gradually on-ramping the new production system before completely decommissioning the old. But there might be scenarios where that process isn't practical, Baer noted.
Improved productivity is the endgame
Oracle's Autonomous Database is expected to free up DBAs to spend more time helping users and developers be more productive with the data in their systems, according to Carl Olofson, research vice president at IDC. It could also free up time needed to evolve databases in a direction that makes sense for enterprise users.
Olofson also expects Oracle's Autonomous Cloud services to bring flexibility to the data side of application development. The self-tuning features mean that when the database schema is changed for developers, the database can be turned over immediately instead of taking weeks to do it. Developers can also set up development and test systems in a fraction of the time typically required for this setup.
To achieve acceptable performance when moving databases and applications to the cloud, users must run their applications in the Oracle Cloud as well, Olofson said. If a company needs to keep some connected applications on premises, he recommended Gen 2 Exadata Cloud at Customer as an alternative. That also requires less staff time because it's managed by Oracle, but IT teams will have to wait for Autonomous Database to become available on it before they can deploy self-tuning and self-managing databases.
Migrating applications and data to Oracle's Autonomous Cloud services is fairly straightforward if the applications are Oracle. "If not," Olofson said, "you need to investigate how to set up your application in the Oracle Cloud and calculate how much it will cost before moving forward."
