Virtualization 2014: Are we still having this conversation?

Virtualization 2014: Are we still having this conversation?

Summary: It's hard to believe we're still having this conversation in 2014, but it's true. There are still companies that haven't made the technological leap to virtualized infrastructure. But there's an easy answer as to why. It's the economy.


I know that you're tired of hearing about the economy. Who isn't? But the truth is that we're all victims of it. The worst of it, from my perspective, was sometime in 2008. It seemed as if the economic momentum had been set adrift. It was the most financially uncomfortable I've ever felt in my entire adult life. But that was then and this is now. If you survived the deep recession of the past few years, then you've emerged on the other side leaner, wiser, and perhaps a bit more conservative. Now is the time to put those lessons learned into action with a "Go to virtual" or a "Go to cloud" plan.

The so-called "Great Recession" prevented a lot of businesses from spending the money to update and modernize their rapidly aging infrastructures. It's not a crime to be fiscally conservative and responsible during those times but if your business is going to remain alive, then you've got to stay competitive by meeting the competition's strategies. Those strategies are lined with virtual servers, cloud solutions, and smarter management.

It is hard to believe, though, that we're still having this conversation in 2014. Virtualization isn't new. Cloud isn't really new either. However, I do realize that there is some fear about diving into the unknown worlds of virtualization and cloud (especially), when you're just now beginning to emerge from muddy waters.

I'm not really one to speak in jargon, particularly business jargon, for very long. Let me give it to you straight up, the way you've come to expect me to do.

If you don't currently have a cloud plan in the works or on the table, then you're behind the pack. 

Here are ten things you need to do now to catch up:

The solution is to first evaluate where you are with your infrastructure. If your hardware is more than five years old or approaching the five-year mark, you need to mark it for decommissioning.

Second, for your hardware that is still good for at least two more years, you need to examine it and find out if it has virtualization capability (it likely does) and what the maximum memory specifications are for it. I recommend that if you have systems that can't use at least 64GB of RAM, you need to use them for test, development, or lab systems. If your systems do not meet the AMD-V or Intel-VT specification, decommission them.

Third, evaluate your network. If you aren't using at least gigabit Ethernet between server systems, you need to upgrade your network hardware, including server NICs.

Fourth, you need to assess your operating system environment. Anything older than Windows Server 2008 R2, needs to be upgraded or decommissioned. Any Linux system older than Red Hat Enterprise 6.1, you guessed it, upgrade or decommission. If you're running Solaris (SunOS), stop that. Switch to Red Hat Enterprise Linux and move forward.

Fifth, select a virtualization host platform. You have several choices: VMware, Citrix XenServer, Hyper-V, Parallels, Proxmox, and Red Hat Enterprise Virtualization. There's really no wrong answer in this realm. You have to select the one that's right for you based on your hardware and on your needs. I suggest starting with a good evaluation of VMware and Hyper-V and going from there. You also can mix your environments, if that fits your needs. In other words, use Proxmox for Linux containers, VMware for heavy lifting, and Hyper-V for some of your Windows servers. Just a suggestion of a place to start.

Sixth, set up a cloud strategy by evaluating some cloud platforms. Again, begin with the big names. Remember that cloud doesn't necessarily mean outsourced infrastructure, although it can. You can develop your own in-house cloud solution. You can also augment that in-house solution with an external, outsourced one to create what's known as a hybrid cloud. I'll cover the advantages of hybrid cloud in another post.

Seventh, think about moving toward a more mobilized workforce by deploying enterprise applications with solutions such as Microsoft's App-V,'s Application Server XG, or Citrix. In these days of mobile phones, tablets, laptops, and Chromebooks, consider moving away from traditional heavy application installations and maintenance.

Eighth, you should look at upgrading your business broadband connection to give your employees a better Internet experience than they can have at home. There's nothing more embarrassing or frustrating than having to make excuses for your Internet connection's slowness or intermittent connectivity issues to a client.

Ninth, take a good look at your backup, restore, and recovery strategy. Virtualization can soften the blow of any hardware mishap. Some vendors now offer hot spare systems to take over the workloads from your business critical servers in case of a failure. Nobody thinks about backups or recovery until something bad happens. It's in your best interest to use the insurance of backup and recovery before you need it. 

You should also look at a cloud-based disaster recovery solution, while you're at it. If all of your hardware is under one roof, you're taking unnecessary risk. 

Tenth, and finally, take a good look at your security. I know that everyone talks about security but you also can't afford to ignore it. A border router with a firewall isn't enough. Requiring strong passwords isn't enough. Antivirus and anti-malware software isn't enough. You need to combine all those things with a serious security strategy for mobile devices, servers, and workstations.

You're putting yourself, your intellectual property, and your finances at risk if you don't have an ongoing security strategy. Virtual and cloud systems aren't any more or less secure than physical ones, so maintain your vigilance and your momentum in providing the best security for your assets.

The movement toward virtualization and cloud are conversations you should have had long ago but if you haven't, now is the time. You have a lot of work to do and you have several options for solutions. If you have questions, feel free to ask me. I'm here to help. You can use the comments section for public questions or you can contact me via the Author Contact form.


Topics: Virtualization, Cloud, Security


Kenneth 'Ken' Hess is a full-time Windows and Linux system administrator with 20 years of experience with Mac, Linux, UNIX, and Windows systems in large multi-data center environments.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Not a big surprise

    For SMBs, virtualized infrastructure isn't as clear cut as the sales guys make it out to be. Sure, you can buy a powerful server, fire up Hyper-V, and run everything virtually. However, if your host server dies, you are dead in the water.... so you need at least two servers capable of acting like hosts to have any kind of realistic redundancy. That isn't cheap for anything but the most trivial of environments. Compare that to individual servers. Sure, you give up the ease of backing up and restoring entire VM images, but if hardware goes down it's ONE SERVER, not the entire infrastructure.

    There is certainly a place for virtualizations, but for companies with less than 20 servers isn't a black and white decision.
    • @croberts

      You're right. That's why companies with those kinds of assets (just a few servers) should seriously consider cloud outsourcing or using virtual private servers. At the price they are these days, you can't beat them.
  • I guess I'll be virtually interested

    and try virtually care
  • Cost is an issue

    We had an issue where the renewal on the SAN would be $5K, yep 5 thousand bucks for an extra 2 years. We decided to buy a new server ($4K) with enough HDD room to move 8 virtual servers (Replicated with Veeam) to the new server in case the SAN goes. So for the same price as the SAN renewal we just create a DR site thats capable of running all our VM's.
    Yes its a risk and yes it will be slow but you just cant keep throwing money at physical storage and server support renewals, for a small business of < 100 people we cannot afford to keep spending huge amounts and if they lowered their prices then we wouldn’t be considering the cloud.
    I think the issue is, places like IBM, HP are charging way too much for their hardware ($10K for one physical host just to run 6 servers with say 64 GIG of ram and $7,500 for a SAN with 12 TB of raid, $1500 for VMware ESXi a year), this is why people still use physical servers (We have 2 just in case, one for backup and the other for hosting critical files if need be).
    I can buy a decent little IBM server for 800 bucks and throw in some 3 TB drives to act as a file server maybe with DNS and DHCP for under $2 grand its up and running.
    So you can how 20 grand is gobbled up very quickly so you need the second host to create fault tolerance for VMware so add another 10 grand to that and its getting close to 30. Then we have to so called support costs per year which you just end up talking to some bloke in India who doesn’t even know what they are doing!.
    Sorry but i still strongly believe in having physical servers and creating my own cloud because if access to the net cannot be reached, the cloud hosted stuff is useless.
    • Cost analysis

      For our clients, we start with the non-IT pieces of their business they can/cannot do without, and the costs of keeping those pieces operating at capacity. For most, the IT infrastructure doesn't generate revenue, it enables other business assets (people or processes) to do so. Could many of them run their entire business on one server? Sure. Can they afford to shut that server down for hours or days due to fire/hack/crash/theft, etc? If they have gobs of cash, sure. If not, the opportunity cost of NOT having certain people or processes available must be considered. That's where risk tolerance, competitive pressure and the value of client/customer/employee good will takes center stage.

      In a dynamic business environment, every piece of the puzzle has a cost, and going without any important piece has a consequence. Deciding what to virtualize, how to do DR and whether warm or hot spares in the cloud are a good investment starts with understanding what people or parts of the process you can do without, and for how long, and what the business will look like when you finally get everything back.

      If you're a gambler, you don't care. If you're a manager, you better.