Virtualization projects have rolled out at a tremendous pace. To keep up with this pace compromises have been made on physical systems that reduce the ROI of the virtualization project. Larger physical hosts are purchased and their compute resources are not adequately provisioned, networks are opened up so that all of the host can get to other hosts with in the virtualized cluster and most compromised is storage resources that are underutilized, wasting TB’s of disk space.

The net effect is that the gains made on the primary driver for server virtualization, increased server utilization, are lost by decreased network and storage utilization.

To turn the tide, increased utilization in all areas and to broaden server virtualization further into the enterprise will require a software tool like Tek-Tools Profiler for VMware that can provide a central view that the server, storage and IT management can use to unify the decision making process throughout the workload lifecycle. Selecting a tool like this can increase the ROI of the virtualized infrastructure, the storage environment and make IT personnel more efficient.

Optimize the Current Environment

With any project, the first step is always to understand the current situation and the first accomplishment that tools like these can bring is the provision of a common viewpoint for all the IT personnel to point to. Tools like these can provide a look at the environment from the angle of storage out to the physical host, from the virtualized cluster down through to storage or even from storage to the virtualized machine.

The result is a holistic view of the virtual environment. One that can be viewed from a storage centric, virtualized centric of physical host point of view.

With an understanding of what is in place the next step is to optimize the current environment. Great gains can be made not only by optimizing the current storage resources but also the currently virtualized servers.

For example some companies use a default template of 100GB per virtual machine since most machines use 20-30GB just for the OS and initial application. The challenge is that for most virtual machines this is too large and then for a few it is too small, seldom is it just right. In a deployment of 20 virtual machines per physical host, this can amount to 2TB’s of allocation per physical host. Consider that even a small VMware environment has 3 to 5 hosts this can amount to 6 to 10TB’s of storage just in this small VMware environment.

Hard Dollar Return on Investment

In many cases an integrated virtualization and storage management tool will expose that well over 80% or 4.8 to 8TB’s of this storage is unused. Keep in mind that most virtual storage infrastructures run on tier one, high performing, highly reliable and expensive storage costing conservatively $12,000 per TB. Identifying this unused storage and reallocating it or at least not purchasing more of it can result in an immediate ROI savings of between $48,000 to $96,000.

Having a tool that will expose this upfront is half the battle. To maximize storage efficiency the IT staff needs to be comfortable in aggressively restricting the storage utilization on a per virtual machine basis. To do this will require that the tool chosen can monitor and trend utilization rates on a real-time basis, warning when capacity is being approached so that allocation of additional capacity can be made. Tools like this can provide instant savings in current storage allocation and defer the purchase of additional storage for potentially years into the future.

In addition an integrated tool can also review the current compute resource allocation in order to maximize the efficiency of the virtual infrastructure. A real-time tool can collect details about each virtual machine to ensure the infrastructure is being leveraged to its maximum while maintaining headroom for the migration of workloads between physical servers. It can also be leveraged to identify the best candidate available to migrate a virtual workload to, in the event of a failure or need for additional compute performance. 

Optimize New Additions

With the current environment optimized, the next step is to further amortize the investment in both the virtual infrastructure and storage by integrating more stand-alone physical systems into the virtual infrastructure. Many virtualization projects grind to a halt after the initial wave of server consolidation. This is caused by the realization of the complexity of managing storage in this environment and by the realization that the remaining servers are not minimal simple workloads. They are increasingly critical to the organization and they have significant compute, network and storage requirements.

Integrated virtualization and storage management tools can provide two solutions to safely move the virtualization project forward. First they can provide a map of what is in place now and show a time elapsed view of virtual and storage environments. That view can be consolidated at a physical server or virtual cluster level for simplicity.

For new workloads the characteristics of the potential server can be examined and a manual but informed decision can be made as to which is the best candidate to receive the newly virtualized workload. Tek-Tools takes this a step further by offering a simulator that will project what the impact of the new workload will mean to the current compute, network and storage resources.

With an integrated monitoring tool, current stand-alone servers can have an agent installed on them so their actual usage parameters are all captured. This includes not only compute, network and storage parameter but also measurements of those parameters over time. This allows for the capture of important trended data to make sure that peak and valley times can be accounted for and a proper virtualization plan can be developed.

Once this data is captured for the correct work cycle then a report can be generated that will identify and rank the potential candidates and match them to the appropriate physical host, including details like sizing of virtual disks which are included in that report.

Manage the Virtual Lifecycle

Lastly as the virtual infrastructure ages an integrated virtualization and storage management tool can identify which virtual machines are either not being utilized at all or not being utilized enough to be justified. Known as virtual machine sprawl, these idle machines are similar to old files in that they consume disk capacity and compute resources. Once indentified, tools exist within most virtualization solutions to archive these machines to less expensive media and then remove the original VM from the infrastructure.

Selecting of a tool that provides an integrated view of the virtualized server environment and its storage can provide immediate ROI benefits by optimizing the current infrastructure, providing the capability to make wiser choices on extending that infrastructure and maintaining it long term.