What I haven't shared yet was my real motivation -- a deep and fundamental shift in my personal perspective of what's really going on in the enterprise IT industry -- and what's happening in enterprise IT shops.
For me, when I realize the world has substantially changed, then I inevitably have to make a change in response.
That's a decent piece of career advice, by the way.
Needless to say, ours is a dynamic industry -- both on the supply side and the demand side.
Enterprise IT shops are getting slammed to do more with less. They don't have the time, the money or the people to address even a small fraction of what they could potentially contribute.
Worse, they tend to spend far too much time on things that don't create unique value, and far too little time on things that could really move the needle.
That's the uncomfortable truth.
Enterprise IT vendors are getting slammed as well: the commoditization of IT infrastructure, death-match competition, overly-funded startups running amok, the cloud in all of its many forms, activist investors -- you name it, it's happening.
That's another uncomfortable truth.
What attracted me to Oracle is that they have good, solid and substantiated answers to both challenges.
It's All About Control Points
The people who pay the bills care greatly about the former, and almost nothing about the latter.
Perhaps the most interesting view of enterprise IT strategy and operations is through the lens of control points -- which roles control what functions, and why?
The Storage-Centric View
All the enterprise's information lives on storage, so -- theoretically speaking -- shouldn't the storage administrator be in a privileged position to make decisions and tradeoffs based on the value of information?
After all -- he or she has complete visibility to every byte of that precious enterprise information -- right?
Unfortunately, that's not the case. A storage admin has precisely zero knowledge of the importance or relevance of information unless he or she is explicitly informed of its business value. After all, from a storage perspective, it's all undifferentiated 1s and 0s.
To gain that knowledge, workflows have to be engineered to explicitly inform the storage team what's important, and what's not. Changes in status take time and effort to propagate; feedback on infrastructure tradeoffs and optimizations back to the people who understand the information's value take even longer.
Not an ideal situation, and no obvious way out of it. Not a winning model, in my view.
The Hypervisor-Centric View
At least the virtualization admin sees all the individual VMs, and can make better informed decisions about information and how it's handled in the infrastructure.
True, but there are limitations here as well. For the vSphere administrator, their abstraction is the virtual machine. It may contain an important application and its data, or an unimportant application, or maybe an application with a mix of both.
The intrinsic value of enterprise information is largely opaque to the virtualization administrator, in much the same way that it's opaque to the storage administrator. Both need to be explicitly informed by someone else, changes are cumbersome to propagate, tradeoffs and optimizations can't easily be discussed, etc. Same story, different actors.
If enterprise IT is all about enterprise information, and making informed choices about how to treat and manage that information, which role is most privileged to make those decisions?
Sorry infrastructure folks, it's the database administrator.
The Database-Centric View
Another uncomfortable truth?
Despite everything you might read in the popular media, that's not changing and that's not likely to change anytime soon.
The DBA sees it all -- applications, middleware, databases -- all the stuff that really matters to the enterprise. They know what's exactly needed: access, performance, protection, security, etc.
One intellectually appealing element of Oracle's infrastructure strategy is simple: empower the database administrator to make informed infrastructure choices from their privileged perspective. If it's important, it lives in a database, and most likely an Oracle database.
Who else would be better privileged to understand the intrinsic value of most critical enterprise information?
That database-privileged view extends to the infrastructure products Oracle builds: Oracle Engineered Systems (Exadata and its brethren), Oracle Application Engineered Storage (the all-flash FS1 and the ZFS Storage Appliance), data protection (RMAN and the new Zero Data Loss Recovery Appliance) and equivalent (and compatible) offerings in the Oracle Cloud.
Each had a database-centric view. Each has a raft of unique and compelling Oracle database-specific features that you just won't find in generic server, storage or virtualization solutions. Each is co-engineered and delivered as a product, not a project.
Coming from my previous background, it's an eye-opening experience.
No one else can do what Oracle does, period.
Here's the equation: if enterprise information matters, the database administrator ends up in a privileged role. Oracle designs engineered infrastructure solutions for this information-centric model.
It's unique, it's differentiated and it creates significant value. And only Oracle can do this.
The Industry View
I've been in and around the infrastructure end of the IT vendor business for over 30 years now. Times are not great for most.
On the other side, public clouds are starting to change the fundamental consumption model for enterprise IT -- infrastructure and applications as a variable operational expense, and not a capital investment.
I think Oracle has great answers to both. Oracle's legendary database and application franchise creates ample opportunity for highly differentiated answers in a world of me-too. The Oracle Cloud is built for enterprise IT -- technologically and operationally compatible with Oracle's on-premises solutions.
Both are showing signs of strong growth. More enterprise IT leaders are coming around to the view that information matters, and productized solutions that are inherently information (and database) centric can offer a better answer than generic, un-differentiated do-it-yourself projects.
The enterprise IT world is changing, and changing fast.
And rapid change can be an uncomfortable -- and unpopular -- thing.
Like this post? Why not subscribe via email?