It’s been a couple of weeks since Oracle once again invaded San Francisco with their 2014 edition of Oracle OpenWorld. Having had a chance to digest and reflect a bit, here are some of my takeaways from this year’s event.
Queue to get into the OOW keynote
Having been to a few of these over the years the format of OOW feels rather familiar (including the queues to get into the opening keynote) and there were no surprises this year either. However the conference format wasn’t the only no-surprise this year. I walked away with two more, as well as two surprises.
No-surprise #1 – Cloud
We had been adequately warned that “cloud” would be the big theme at this year’s OOW, and in deed it was. Both in a real sense in that Oracle actually did announce new cloud services, but also in a sense that the term cloud was incorporated in a lot places that had little to do with cloud.
The real news was actually the introduction of Oracle Cloud PaaS services for database and middleware, as well as IaaS services where you can rent a VM with Oracle Linux on it. Oracle is late in the game when it comes to cloud. Really late. Amazon have been doing IaaS for decade, Microsoft for a few years. For developers who wants PaaS it’s been available from Microsoft, Google, Salesforce and others for 5+ years. Even those who particularly want an Oracle database in the cloud have been able to get that from Amazon for years and from Microsoft since last year. Hence I don’t see Oracle Cloud succeeding as a generic IaaS / PaaS platform competing on price, breadth of technologies supported, and global data center build-out.
When looking at the initial PaaS services and their pricing it seems that Oracle indeed has a different target market in mind. The value proposition of Oracle cloud is not price, nor global availability. Instead the unique selling point is that correctly configuring and managing a database (Oracle or any other) for extreme availability and performance is complex and requires specialist skills. A requirement few organizations are equipped to handle. At $1,500 – 3,000 per month, per CPU core, Oracle Database-as-a-Service isn’t cheap. But neither is licensing the software and staffing a 24×7 team of highly skilled Oracle DBAs.
In the long run I think we could see some large customer with Oracle based systems migrating to Oracle cloud. But it will take some time. The data centers must be built out (upon launch the PaaS services were only provided out of one US data center), support for non-Oracle OS and products is needed since few applications are pure-Oracle, options for connecting a customer’s on premise IT in hybrid scenarios need to be provided.
No-surprise #2 – In Memory
Although announced at last year’s OOW, Oracle In-Memory option wasn’t actually released to the market until a couple of months ago. So it wasn’t a big surprise that the new in-memory technology got a very prominent place also this year. Both to educate developers and partners, as well moving Oracles positions on in-memory forward in light of all the attention SAP have been getting with Hana.
When it comes to in-memory technology I find myself agreeing with Oracles approach. To me it makes no sense for the whole IT industry to re-learn and move to a new purist in-memory database, when we can have all the same benefits built into the databases we know and use since decades. It didn’t make sense with XML databases, it didn’t with Object databases, and it doesn’t with in-memory databases.
A bit more interestingly Oracle is taking a tangent approach also with Big Data analytics, integrating Oracle Database with various big data technologies and extending the use of SQL to apply also to big data.
Surprise #1 – Applications
One thing that did surprise me this year was the lack of major news around Oracle applications. No major new Fusion Applications were announced (there is still no Oracle Fusion Manufacturing, for example). There hasn’t been any major release of eBusiness Suite for a couple of years, and from what I picked up there won’t be for the next few years either. Sure there are some smaller functional enhancements, the UI gets updated in spots (but most of it not at all), and the odd mobile application for casual users. But no major release, no architectural changes, and nothing of that mentioned in the roadmaps.
I’m not sure what this means—either Oracle have prioritized other things (building out the cloud, enabling their applications to run in the cloud), or there is something else coming soon and thus Oracle is ramping down investments in current applications.
Surprise #2 – Larry
With Larry stepping down as Oracle CEO just weeks before the event I really wasn’t expecting to see him get much, if any, time on the main stage. But it turned out I was very wrong on this one. Larry did the opening Sunday keynote and came back on the main stage a couple of more times throughout the event. I think he got more time in the spotlight than new CEOs Mark Hurd and Safra Catz combined. He apologized for his no-show last year, and even managed to walk through a couple of simple demo scripts.
Predicting OOW 2016
Always dangerous to make predictions, but for what it’s worth I predict that at OOW 2016 Oracle will lead with announcements of Oracle applications, and do a re-launch of Oracle Cloud (like they did with in-memory this year), which by then should feel a bit more mature
