Starvation by virtual invisibility

Without a doubt, the cyclical world economy is making for an interesting workplace. It looks like we're destined to go from eating off the fattened milk cow while the economy is rolling, to picking our teeth with the bones we have left over during the famine times. Big picture, that's not all that unusual right? History's filled with cyclical feast and famine events, and we've just taken it up a notch, so now it plays out on corporate ledgers and in the stock market.

One interesting offshoot of this last cycle is the tremendous uptake of server virtualization while we were feasting. Without talking about the why, its obvious there's a good few virtualized server installations out there in businesses small and large. What's interesting, is wondering what type of shape they're in going into what increasingly looks to be a recessionary famine. Is you're virtual infrastructure set for the long haul? Do you have the management and security tools to make your virtual servers be everything they can be? From my perspective, most folks fall far short of their ideal. The problem is, most of these folks are thinking they need to cinch their belts tighter, and in turn don't have a lot of financial ability to do anything about it. I'll leave the milk cow analogy behind for the moment before it gets distasteful, but all is not without hope - there is enough emerging capability in the marketplace that you might just find good solutions that can help you stretch your virtual infrastructure for new needs. How's this? Well, lots of us shy away from certain virtual infrastructure use cases because of lack of control - maybe PCI compliance because we can't manage security, or maybe critical applications or heavy workloads because we can't holistically and granularly manage resources in the virtual infrastructure.

The problem is, when it comes to virtual infrastructure management, there's lots of confusion in the marketplace, and a lot of solutions that do different parts of what is a complex job. Complete virtual infrastructure management is about more than lifecycle management, and more than automating tasks like change auditing, configuration management, provisioning, etc. I recently talked to one vendor that has applied an interesting set of capabilities to virtual infrastructure management and, by attacking it from a unique angle, has a bit more to show for their efforts - Reflex Security.

The thing is, Reflex Security is about a lot more than security. Once upon a time, along with dozens of other security companies, they crawled up the operating and application system stack to understand events at the application layer and deliver an intrusion detection and security solution. But for a good while since then, they've been leveraging their application level understanding to dive back down the stack, and provide management and visibility for the subcomponents of a virtual infrastructure - servers, switches, adapters, and more. What they have is a huge correlation engine, where all elements of the virtual infrastructure can be tied together - if your performance is off, look at where your apps, servers, and switches have changed, historically. Just as easily, you can still employ all of Reflex's traditional security and control features.

I've mentioned before that I see a new generation of information management emerging that is context-based, and in many ways, I see Reflex delivering this same level of context around the virtual infrastructure. Context becomes more important the bigger your infrastructure is, and becomes a major hurdle in the cloud (more on that in a later post).

In a way, a timely strategic purchase of the right infrastructure management tools may be like superfeed for your virtual infrastructure, and get you through the famine. But you have to figure out what makes your virtual infrastructure all it can be, without throwing more management overhead at the problem, or creating an infrastructure that is too complex to align with your needs (don't accidentally turn your calf into chicken feed, that won't keep you going either).

In my opinion, where many solutions fall short is a limited perspective approach to management - network vendors attack the problem from a host IP perspective, vendors that are heavily weighted toward some set of applications may not be broad reaching enough at the application level, business process folks often dive off into strictly process automation, and dashboards and data collectors may inundate you with data while giving you little visibility into what is really going on in your virtual infrastructure. Worse yet, everything gets muddied up even further because the average administrator using VMware often can't tell where VMware's Virtual Infrastructure tools end and third party tools begin. In my opinion, if you're a server virtualization owner, and looking to feed your calf, you better pay careful attention to this space, and look for the full picture - that means a deep ability to tie what is happening at logical levels to the physical level, and then take action, or otherwise support your management processes, based on those relationships and events.

Final thoughts: This famine analogy comes to an interesting end. In the past, you used to starve first, then become virtually invisible. Today, if you have virtual invisibility, you may starve from your inability to manage it.