OK, I’m a sucker for cool words.
Especially when they neatly summarize a very complex and nuanced discussion. I heard this one being used at one of our internal VMware meetings, and it rang like a bell in my brain. I’ll reserve the right to give credit where credit is due when I track down the original source.
Hybrid — as in hybrid clouds. “-icity” — a measure of degree, as in “elasticity”.
Hybridicity: a measure of just how hybrid is a particular cloud computing environment. In this context, more hybridicity is better than less hybridicity.
If we’re going to see more mainstream adoption of public cloud services by enterprise IT groups, I believe the core argument is going to be around hybridicity — how much, what kind, how good?
Standard Experiences Matter
I travel a lot, and I frequently rent a car. Fortunately for me, driving a car rented from vendor A is remarkably similar to driving a car rented from vendor B. The control surfaces are the same. The behaviors are the same. And, if something should go wrong, my problem resolution process is the same.
Not to belabor the point, but imagine if every car rental agency rented fundamentally different cars?
This one uses a familiar steering wheel, but this other one uses handlebars. The throttle on this one is a pedal, but over here it’s a squeeze grip. The fuel gauge is easy to read here, but this other car rental company put it somewhere I can’t see, and I’m not sure what it’s telling me.
Silly, right? Of course all cars use a consistent interface with consistent behaviors, otherwise how could you expect people to rent one?
But that’s exactly the situation we’re in when it comes to today's public clouds.
Every vendor has “their way” of doing things when you go to rent infrastructure. Maybe it’s better, maybe it’s not — but it’s guaranteed to be not only inconsistent with other public clouds, but also inconsistent with what most enterprise IT groups are doing internally.
And people wonder why public cloud adoption is so low in mainstream enterprise IT — there is very little in the way of a consistent experience.
Or, put differently, very little hybridicity.
The Reality Of Enterprise IT
I’ll let you in on a big secret: enterprise IT really isn’t about technology so much, it’s about people and process. Look on any IT "balance sheet", and you'll quickly see the massive investment is in people: their skills, their knowledge and how they go about doing their work.
The best technology in the world won’t fix people and process issues.
Present these people with a technology, or a public cloud service, that is incompatible with existing skills, tools and processes, and it immediately becomes an unattractive proposition.
The barrier to adoption is quite high, and IT management is faced with the additional requirement to peanut-butter the available skills and resources even thinner to cover the new, inconsistent thing.
Keep in mind, the argument here is not that the public cloud service is theoretically “better” or not, just that it’s quite often inconsistent and incompatible with what an average enterprise IT shop is doing today.
The Goal State Of Hybridicity
Let’s take an average enterprise shop running, say, a moderate number of VMs, maybe a thousand or more?
Ideally, any public cloud resource would appear as a simple extension of their existing clusters.
The behaviors would be the same. The control surfaces would be the same. Troubleshooting problems would be the same. Things like load balancing, and setting up for high availability and DR would be the same. Monitoring and reporting would be the same.
You’d have to look hard for differences between the on-premise resources and the rented resources.
Just like jumping into a rental car at the airport, you'd be able to figure it out without too much trouble — and start driving immediately. And switch to another provider with a minimum of hassle.
But that’s not the world we’re in — why?
Needed: A Consistent Starting Point
The norm in many enterprise IT environments is home-grown: a kaleidoscopic array of tools and scripts, a bewildering set of processes — some ad-hoc, some more formal. While that’s not ideal on several levels, it is what it is.
If we put up the concept of hybridicity and public clouds against this unpleasant reality, the challenge is obvious: by definition, there can’t be a compatible public cloud service that runs the way the average enterprise IT shop does -- they're mostly one-offs.
How can a public cloud service expect be compatible with each and every enterprise IT snowflake out there?
I’m not going to pin the blame for low public cloud adoption by mainstream enterprise IT on this and this alone. There are the unarguable economic advantages of owning vs. renting an asset if it’s highly utilized, for example.
Here is where I am optimistic.
I have now met enough enterprise IT shops heavily invested in VMware’s vCloud Suite and vRealize Suite to observe a pattern: they all have a reasonably consistent set of skills and operational processes — not to mention identical technology choices in key areas.
Much like SAP implies a good way to run your business, vCloud Suite and vRealize Suite imply a good way to run enterprise IT: continuous service delivery, consumption portals, cost transparency, integrated management and monitoring, etc.
These tools encourage processes and methodologies that make sense, and are inherently consistent with other similar users.
Here’s the win: these outfits as a group can now consider a compatible public cloud service (vCloud Air, or more broadly the vCloud Air Network) without facing the cliff of learning how to drive a car all over again. Other considerations (economic, etc.) are still in play, but a very substantial one is now off the table.
As such, this group enjoys an advantage that others don’t — easier and more productive consumption of compatible public cloud resources, should they be interested.
By standardizing on these particular tools and workflows, they’ve achieved a two-fer: a better operational environment, and new convenient consumption options via hybrid cloud.
I also get exposed to organizations trying to figure out their management stack going forward. All sorts of mix-and-match options out there. Only a few of them consider whether or not there will be compatible public cloud consumption options as a result of their choices.
Maybe they should.
Back To Hybridicity
When it comes to complicated things in our lives, we demand standardization. Cars work a certain way, home entertainment systems work a certain way, computers, smart phones, etc. We value consistency in the user experience, the control surfaces, the underlying behaviors — and what to do if there’s a problem.
Enterprise IT is no different. A stupendous amount of effort is invested in people and process which results in a de-facto standard for each and every IT organization.
Moving that investment forward seems to be job #1 for most of the IT leaders I meet these days. As it should be. “Getting to cloud” really means two things: transforming internal IT operations to an as-a-service model, and establishing convenient consumption options that are entirely compatible with their skills and operational processes.
Hybrid cloud, in its essential form.
And it seems that there are more than a few VMware customers who’ve cracked this code.
-------------
Like this post? Why not subscribe via email?
Comments