Reminder - Oracle 18c is autonomous only if Oracle runs it for you

Profile picture for user pwainewright By Phil Wainewright January 29, 2018
Summary:
People need reminding - the new Oracle 18c database is autonomous only when Oracle runs it for you. That will drive adoption of Oracle IaaS, PaaS and SaaS

Oracle autonomous cloud availability
This shouldn’t be news to anyone, but a lot of people haven’t figured it out yet. Oracle 18c marks a turning point for Oracle’s flagship database. Autonomous operation — its most distinctive feature — won’t be available if you choose to install and run the product yourself.

Oracle 18c is not only cloud-first, it's cloud by preference.

This reflects a crucial shift in Oracle's positioning. It's no surprise therefore that it's taking time to sink in. But once people start to catch on, the effect on Oracle's business will be significant. First of all, it rapidly accelerates the rate at which enterprise customers make their move to the cloud. And secondly, it significantly lowers Oracle's own costs of serving modern cloud applications to its customers. Let's examine each of these in turn.

Autonomous isn't a shippable feature

One of the most important points to note about the autonomous messaging is that the underlying technology hasn't changed all that much. Despite the big leap forward in product numbering, from 12c to 18c, the product itself is changing only incrementally, effectively from release 12.2.0.1 to 12.2.0.2. Switching the name to 18c reflects the change to an annual release cycle, and so the new number simply catches up to the calendar year.

The reality is that 18c is the culmination of incremental technology evolution over many years, including Real Application Clusters (RAC), Oracle 12c's multi-tenant container architecture, and other more recent additions.

The one big change with 18c is autonomous operation, which is not a shippable feature. While 12c is already a 'cloud-first' product in the sense that successive releases have become available on Oracle Cloud before shipping to customers, customers have always had the option of waiting for general availability to install it themselves. That's not an option for the self-driving database.

Autonomous operation is cloud-only, and only on Oracle Cloud. Anyone who buys Oracle 18c to install and run on Amazon Web Services — or in their own data center — will be sorely disappointed if they expect it to repair and tune itself without manual intervention. That's only available when Oracle manages the database for you, either on Oracle Cloud Infrastructure (OCI) or as part of its Cloud at Customer service, a remotely managed instance that runs in the customer's data center.

Theoretically, it's possible a customer could tap into the same features within Oracle 18c to attempt their own version of autonomous operation. But they would lack two crucial components. First of all, they can never have the same visibility as Oracle's own operations teams into how this proprietary software is engineered.

More importantly, they'll never have the breadth of data that Oracle has gleaned over many years of running thousands of customers' databases in the cloud. Therefore any machine learning model they could build to inform their autonomous operation would inevitably be inferior to Oracle's. No doubt a few stubborn souls will attempt it nevertheless, but it's a fool's errand.

A fat carrot and a big stick

So to get the benefit of autonomous operation, enterprise customers must migrate their database instances to the Oracle Cloud. To encourage them, there's both a fat carrot and a big stick.

The fat carrot is lower cost. When Oracle presented 18c at last year's OpenWorld conference, a lot of the focus was on the cost differential versus running the same database on Amazon infrastructure. The point was not to bash the competition — though that's always fun — but to ram home the cost advantage. Cost arbitrage has always been a primary reason why people move to the cloud. Now Oracle offers its customers certainty that the cheapest way to move their Oracle database to the cloud is to move it to the Oracle Cloud — guaranteed, says Oracle, to be 50% cheaper than running it on AWS.

The big stick is the ever-present risk that customers face when they fall behind on patching their on-premise database estate. Enterprise CEOs are increasingly conscious of this risk, according to Oracle CEO Mark Hurd:

Most of our customers take, I don't know, nine to twelve months to patch all of their databases, after we release a patch to our database. That problem — just that problem, no other problems — just resolving that problem changes our customers’ entire business. This is a database that self-autopatches.

If it did nothing but solve that problem, and that entire user base, tens of millions of cohorts, move to take advantage of that capability, the world changes.

C-suite pushes enterprise to the cloud

Speed of course is relative when it comes to enterprise IT — these migrations are highly complex operations. As an example, retailer The Gap expects to take 18-24 months just to complete its transformation program to public cloud, having already spent more than a year in research and planning. Even then, the team had not yet, when we spoke to them earlier this month, evaluated the potential impact of autonomous operation.

That's a fairly commonplace scenario. IT teams are looking at what they've got and how to move it — much of it still on earlier versions than 12c. What's available once they've reached the cloud is something they'll think about when they get there.

This is why Oracle's Hurd is pitching his message at CEOs. Left to their own devices, IT will proceed at their own pace. But if enterprise leaders are setting the agenda, IT will get a mandate and the resources to accelerate their progress. The prize of eliminating the exposure of unpatched databases, as well as being able to build on the latest technology without having to struggle through endless upgrade cycles, is persuasive enough for business leaders that want to stay competitive in a fast-moving digital world.

This is a stance echoed by Andy Mendelsohn, EVP Oracle Database Server Technologies, who says the adoption dynamics will be different than in previous generations of technology:

It's going to be a little different than in the past. There's going to be a top-down executive push to lower cost by moving to cloud ... It's going to be a matter of corporate survival — 'Either we do this or we're dead.'

We want to make sure if and when they move to cloud, we're there for them.

The acceleration to the cloud may not look like much to start with — Hurd himself talks of a 5-10 year transformation — but if it's driven by the C-suite, it won't take long for momentum to build up.

Cloud applications go first

One factor that will add momentum is Oracle's own migration of its SaaS applications to its IaaS platform, Oracle Cloud Infrastructure (OCI). This is one of the hidden benefits for Oracle of going autonomous, among others.

The schedule for bringing its IaaS offerings autonomous is fairly well known. The data warehouse cloud goes first, as early as this month. Next, around halfway through the year, comes the OLTP cloud — this is the core, mission-critical database product. Then later in the year there will be an autonomous NoSQL cloud and finally a graph analytics cloud.

What is less well known is that Oracle has an aggressive timetable for moving all of its PaaS services and the bulk of its SaaS applications to its cloud infrastructure, where they will benefit from autonomous operation as well as the many other features of the underlying database technology.

Like many of its customers with their own applications, Oracle has not yet completed moving all of its SaaS and PaaS properties to OCI. According to Oracle President Thomas Kurian, that migration will be completed this year. PaaS services will go first — two thirds are already there, the rest "will be done in February."

That then paves the way for the SaaS portfolio to follow suit. Not all of it, but enough to account for the bulk of Oracle's SaaS revenue, says Kurian:

75% of the SaaS revenue for the company is from a set of products. They will run on top of the infrastructure between May and August of this year. So there’s an aggressive program to migrate our platform and our software as a service on top of this infrastructure layer.

What's the bottom line impact?

Learning this made me curious what the effect might be on Oracle's bottom line, given the company's claims for the cost advantages of running on autonomous infrastructure. I had the chance to put this question to CEO Hurd, and he gave an interesting reply.

While the SaaS products currently run on Exadata servers, they're not necessarily running on the latest version of the Oracle database — some are still on 11, others may be on 12 but are not necessarily taking advantage of its multi-tenant features. Moving to OCI will bring them up-to-date with those capabilities, as well as giving them direct access to PaaS services, in particular AI and machine learning. So it's not only an operational and financial improvement, it also allows the SaaS applications to harness new capabilities within the platform. As he explains:

It isn't just the autonomous database, it's the new features that have come with 12, combined with the autonomous database. Then you can imagine, you’ve got all of this AI feeding all of these applications. The ability now to autopatch, autotune, drive everything on a common, this new set of performance capability that we've got.

It isn't just that it gives us more margin. It will. But it's also the sort of capabilities that it brings with it that we can take to the market that frankly nobody else can do.

My take

This is one of those 'turning an oil tanker' tasks. For its customer base to embrace the cloud, Oracle has to change mindsets that are deeply ingrained in the traditional way of doing things. That's why Hurd has been talking about fundamentally transforming how people approach enterprise IT. It really is that big a change, and it will take years to get everyone on board.

Meanwhile, Oracle's own transformation illustrates how much time this takes, with a significant slice of its own products only now ready to start migrating to its most up-to-date cloud infrastructure. Once that task is complete, however, by the end of the summer, I expect that in itself will add an additional spurt in customer adoption.

Looking back over the history of cloud computing, even while a lot of the industry focuses on the infrastructure layer, it's always been at the application layer where enterprise spending has been predominant. Once the decision has been made to migrate to the cloud, it often makes more sense to adopt a ready-made SaaS application rather than attempt to migrate an existing application to a cloud platform. So don't be surprised if the biggest impact of the autonomous database, paradoxically, turns out to be a surge in customer adoption of Oracle SaaS applications.