Buzzwords. The tech industry is inundated with them. Big data, “disruptive” technology, software-defined X, hyper-convergence, automation, DevOps, and of course The Cloud. I’m not exactly sure what it is about the term “Cloud Computing” itself, but saying it aloud seems to make many of the individuals that I engage with who manage IT infrastructures cringe.
Perhaps it’s due to the fact that it appears on the internet over 120 million times. Or maybe it stems from the fact that most Americans are flat out confused by the concept of Cloud Computing according to a 2012 national survey. Whatever it is, there seems to be a growing combination of distaste and/or confusion surrounding the term ever since Compaq allegedly coined it in late 1996. After all, Cloud Computing is just another form of virtualization isn’t it? Well, not exactly.
To make matters worse, underneath Cloud Computing itself we have an expanding glossary of other terms and buzzwords that include private cloud, hybrid cloud, cloud bursting, SaaS, IaaS, multi-tenancy, and cloud automation.
Stay with me now, but even cloud automation has its own subset of terms ranging from automated configuration management to automated self-service and provisioning. But as a colleague of mine referred to in an earlier post, what are we really trying to automate?
Does any of this so called automation help us truly control our continuously growing data centers and cloud environments?
The first sentence of a quick Wikipedia of “automation” reads as follows: “Automation or automatic control, is the use of various control systems for operating equipment…” Wait a minute. Automatic control? The use of various control systems? To me it would seem that a lot of the buzzwords surrounding automation in the IT industry refer to process automation rather than automatic control.
If you take a step back and think about it, all of this automation pertaining to configuration, orchestration, and provisioning would seem to only complicate things even more when it comes to actually maintaining the environment in a healthy state.
It’s All About The Decision(s)
There are numerous articles across VMTurbo.com that speak to the importance of the decision and the complexities around making the correct ones, at the correct times.
Which workload do I move? Why do I move this workload? Do I also need to size this workload? What about capacity?
Well, now let’s think about the complexity of making the right decision at the right time as it pertains to a private and or hybrid cloud. The decisions may be similar, but I’d argue that timing may be even more crucial as I have provisioning, orchestration, and configuration all taking place in an automatic fashion behind the scenes. Leveraging a cloud infrastructure certainly gives me more opportunities in regards to making decisions to maintain performance and efficiency, but it also increases the difficulty of making them.
Maybe I move this workload to another silo? Does the VM need more resources? Perhaps I size the entire pool? When and where will I need additional physical capacity?
Just as with native virtualization, not only is making decisions in real-time as it relates to performance and efficiency extremely difficult, it’s simply not possible. Not without software anyway. And as organizations continue to scale out their private/hybrid cloud infrastructures, in conjunction with automating various cloud processes, the problem only exacerbates.
By allowing software (see VMTurbo J) to offload a lot of these decisions that we are forced to make manually every day, we can alleviate some of the pain. Want to automatically resize pools as end-user demand fluctuates to reclaim idle resources and redeploy them to other users? Let software do it. Want to accommodate various resource spikes by utilizing capacity in other silos? Let software do it.
By letting software (i.e. VMTurbo) automate and CONTROL other software (e.g. a hypervisor) we can rapidly scale not only our cloud environments, but our businesses that are being supported by them as well.