Oracle 18c: Oracle Autonomous Database Cloud

Larry Ellison, Chairman of the Board and CTO, Oracle, announced the launch of the world's first self-driving and autonomous database cloud at the opening keynote of the Oracle Openworld 2017, in early October. Oracle's new Autonomous Database Cloud (part of Oracle 18c) optimistically aims to automate most of the manual (human) processes for upgrading and maintaining a database system.

There has been a lot of excitement due to its AI (Artificial Intelligence) integration and yet there has also been fear and confusion with the advent of the autonomous capabilities. Primarily, it is the Database Administrator community which fears becoming redundant with a lot of the tasks they did manually for years, now being replaced with AI. This article aims at providing a basic insight to clear some of the confusion and hopefully help people set expectations.

First and foremost, the autonomous and self-driving capabilities are limited to the cloud version of Oracle 18c. Which means, all non cloud based setups will continue to require manual/human intervention by DBAs as before.

What is so special about the Oracle Autonomous Database Cloud?
The Oracle Autonomous Database Cloud removes the human labor associated with tuning, patching, updating and maintaining the database. It employs machine learning for

• Performing continuous adaptive performance tuning
• Automatically upgrading and patching the database, while running
• Automatically applying security patches, while running, to protect against cyber-attacks
• Automatically resizing compute and storage, for scaling, without any downtime

As per Oracle, the cost savings will be multiplied because Oracle Autonomous Database Cloud consumes less compute and storage than Amazon, with lower manual administration costs. It can handle different workload styles, including, transactions, mixed workloads, data warehouses, graph analytics, departmental applications, document stores and IoT (Internet of Things).

"This is the most important thing we have done in a long, long time", Ellison said at the launch. " The automation does everything. We can guarantee availability of 99.995%, less than 30 minutes of planned or unplanned downtime."

Oracle also announced the Oracle Autonomous Data Warehouse Cloud, with the same machine learning backbone as the Oracle Autonomous Database Cloud. It is aimed at providing simplicity in running workloads without the need for any manual tuning and will support self-scaling as well.

Also included in the Autonomous suite is an enhanced cyber-security system, which also uses machine learning and is well on the way to becoming fully automated.

"Cyber-security system detects threats and passes along. Database system has to patch itself and not wait for a human to schedule a downtime to gracefully implement a patch in a month or two. These were developed together and designed to prevent data theft. We do everything we can to avoid human intervention" said Ellison.

Another aspect of the autonomous setup, included in the USP, is the simplicity of the whole setup despite sounding like a high tech futuristic technology.

As per the release notes, "Unlike traditional cloud services with complex, manual configurations that require a database expert to specify data distribution keys and sort keys, build indexes, reorganize data or adjust compression, Oracle Autonomous Data Warehouse Cloud is a simple 'load-and-go' service. Users specify tables,  load data and then run their workloads in a matter of seconds, no manual tuning is needed"

"Every organisation is trying to leverage the overwhelming amount of data generated in our digital economy. Oracle Autonomous Data Warehouse Cloud is designed to deliver industry-leading database technology performance with unmatched flexibility, enterprise scale and simplicity. The intent is to ensure that businesses get more value from their data and modernize how data is managed", said Carl Olofson, Research Vice President, IDC (International Data Corporation).

So does this mean the death of the DBA?
This is a legitimate fear, which shouldn't be dismissed outright, but it would better described as moving towards an evolved DBAs job. DBAs will still be vital to large and complex databases, but their role will shift from an operational one to that which requires more thought, planning, analysis and of course development.

Ellison said that more automation won't cost database administrator jobs, but in fact free up those professionals for high level work like security. "A freedom from drudgery. There has been more and more automation in every release for decades. You've lived through the last few decades, you'll live through the next decade also."

The aim of this suite of autonomous databases is to automate as much as is possible to reduce the need for human interaction and free the DBAs from mundane, so they spend less time on infrastructure, patching, upgrading, ensuring availability, tuning and instead use more time for productive tasks like database design, data analytics, data policies and securing data.

There is also the million dollar question of how long it will take the companies to move to an AI based infrastructure. Some industries are still wary even of the cloud itself and many more are unlikely to hand over control of mission control systems to third party vendors any time soon. There is concern about giving up control... Organisations like banks, which are highly regulated, would be the very last to surrender control.

The analytical and development DBAs have nothing to worry about. They will always be needed. The ones that follow a preset recipe and perform monotonous operational tasks are the ones with a possible risk. Then what's the secret to keeping your job?

Never stop learning, keep upgrading your knowledge and your job will always be safe!!

We all know what happened to Kodak for being complacent and not evolving with the times.

Whether this will trigger a race for bringing AI integrated products in the market, or, will the industry wait and watch how the world responds to AI, remains to be seen. There is, however, no doubt, that this is a paradigm shift in the technological priorities of the industry and some exciting times lie ahead...

Oracle 18c is available on-premise, in Oracle's public cloud and customer cloud flavors. The data warehouse version of Oracle 18c will be  available in December 2017 while the OLTP version will be available in June 2017.