This year VMworld tackles what happens when your VM farm grows up and spawns hundreds or thousands of baby VMs
This year VMworld tackles what happens when your VM (virtual machine) farm grows up and spawns hundreds or thousands of baby VMs. It used to be you fired up a bunch of VMs on a server (or your laptop) and that was it. The networking, storage, and perimeter of your new VM world were in one box. But what if you suddenly (or really over the years) have VM creep – where you find yourself the proud parent, er, administrator, of thousands of tiny VMs?
You go to VMworld and try to figure out what everyone else is doing. Lots of case studies by huge giant organizations are de rigueur here in SFO, but most of the thousands of feet attending belong to somewhat smaller enterprises trying to figure out what they have unwittingly wrought.
When your VM crib expands beyond the first box, you have to figure out storage, networking, management and security.
Storage is non-trivial because in an (almost) always-on environment, it’s hard to find downtime anymore. Your clients are logging in at midnight to check emails and stuff like that. So if there was a magic button that just backed stuff up without crashing your clients or taking outages that would be a grand button indeed. But what if that button had to push the backup across state (or country) lines to your disaster recovery site in something close to a plausible timeframe?
Now you have to work on your network. Buying (or renting) big fiber pipes becomes the new hurdle. Latency, especially in years back, was the death of VM performance across the enterprise. It wasn’t just the raw capacity, but ping times killed your virtual desktop apps at the same time they killed backups. Now you need both.
Then someone (or really a bunch of someones) invented Software Defined Networking (SDN). What, you don’t have a SDN sitting around yet? You’re not alone. But if you want to spin up a department suddenly, you might want to spin up a mini-network too inside your other larger networks. That means SDN, or lots of hardware networking folks to build lots of routers. Want to know how that all works? While the technology is still somewhat nascent, the big brains are here, fleshing it out.
So now you have to overcome VM sprawl – that feeling that you no longer really understand with any specificity exactly WHERE your VMs really are, nevermind figuring out if they’re all secure. Sure, you could forklift a bunch of security boxes out across your datacenters and back offices, but even that’s non-trivial, especially if you want them all to play nicely together and give you a high-level view that can explain to an exec. Getting to a single pane of glass that can congeal and express your sprawling organization’s security report card is a super non-trivial feat. Whichever vendor can put that together wins the gold prize.
In the meantime, at least get some kind of a security foothold, even if the centralized management Holy Grail is still somewhat elusive. You have to start somewhere, even if it’s just a trip to SFO to find out what everyone else is doing at VMworld.
If you are at VMworld and want to learn more about security in virtualized environments, including a demo of ESET’s shared local cache solutions, stop by our ESET VMworld booth #2041!