A virtualization deployment offers many benefits, as everyone knows by now. But after the initial consolidation phase, it's common for a virtualization deployment to slow down, or even come to a complete stop.
The reasons behind this phenomenon, called VM stall, are varied. Sometimes it's too challenging -- either technically or economically -- for an organization to virtualize larger, more mission-critical applications. Frequently, bureaucracy and politics get in the way, because these applications have more stakeholders and affect more users. And then there's ISV stall, which happens when application vendors just won't support their products in virtual environments.
Luckily, there are just as many solutions for fixing a stalled virtualization deployment as there are causes. Members of our Server Virtualization Advisory Board share some of their tips as they answer this question:
How can an IT administrator jump-start a stalled virtualization deployment?
CJ Metz, First American
In order to jump-start your stalled deployment, simply look back to why you went virtual to begin with. Virtualization offers a whole host of features that are simply unavailable in a physical environment: high availability, VMotion, centralized management, and the ability to run multiple OSes and applications on one physical server. When you take these things into consideration, the barriers begin to dissolve away.
If your organization is having more bureaucratic issues, the people who are holding things up need to be educated. Moving to a virtual infrastructure is not cheap, and if you are experiencing server stall, letting your expensive infrastructure sit is not benefiting anyone.
Bottom line, people need to understand that virtualization is the way of the future. Mission-critical applications can and do run virtualized. In most instances those same apps run just as well if not better than on a physical system. Assemble the appropriate people and show them why the decision to virtualize is the right one to make.
Eric Siebert, Boston Market
Let's look at the question from three different angles:
Technology issues: VM stall can happen because of improper planning or poor architecture design. Having a proper architecture to meet your workload needs is absolutely critical in a virtual environment. You should bring someone in who is experienced in this area and can help you make the right design decisions and get you back on track.
Bureaucracy: This one can be difficult to overcome. Point out how virtualization could have prevented a certain problem or made it easier to deal with. Often, when a major issue occurs, companies will do anything to prevent it from happening again. If you try selling a sprinkler system to someone who has never had a fire, they will probably be hesitant to buy one. If you try and sell them one right after they have had a fire, they will probably be eager to buy one.
You can also try winning colleagues over by demonstrating the technology to them. Seeing the cool features that virtualization offers in action can help turn a skeptic into a believer.
ISV support: There is a lot of competition in the ISV world. Tell your vendors that you are virtualizing and you expect them to support their products or you will be looking at dropping them and going to their competitors. Virtualization is not a fad, and an ISV that does not support virtualization probably won't have many customers.
If changing vendors is not an option, attach a LUN directly to your VM so you can easily move it to a physical server if needed to resolve a problem.
Shannon Snowden, New Age Technologies
For the last year, our company has been virtualizing servers running one of the most mission-critical applications in the healthcare industry. It wasn't very long ago that the vendor didn't support this application on a virtual platform, but the vendor had to start supporting virtualization because of repeated demands from clients. What sparked the increased demand was the flexibility for disaster recovery that virtualization enables.
Disaster recovery in organizations, particularly in healthcare, is a critical concern. When comparing traditional disaster recovery methods to the disaster recovery capability virtualization offers, it's like bringing a cumbersome, antiquated knife to a gunfight. Many of the vendors who are holding out on virtualization support make mission-critical applications, and because of that, the market will eventually force them to support their applications on a virtual platform.
So how do you jumpstart the next virtualization phase? Focus on disaster recovery.
Dave Sobel, Evolve Technologies
The easiest way to jump-start a stalled project is to transfer responsibility of the stalled piece. When a vendor is refusing to claim responsibility for support, if that responsibility for support can be moved to the internal IT group or a third-party support organization, the obstacle is removed. If a key business decision is holding up the process, finding a higher-level champion for the project will open up the project. The key is being clear about the intention and goal of the project, and finding the right parties willing to assume responsibility. If that can be done at all stages -- and be well documented -- the project will be a success.
Have a question for the Server Virtualization Advisory Board? Email Colin Steele, Site Editor.