Tuesday, November 17, 2009

Storage Switzerland was briefed recently on DynamicOps’ Virtual Resource Manager (VRM) solution for VM lifecycle management. VRM creates a simplified process for the deployment, provisioning and maintenance of virtual machines. At its heart is a workflow engine that organizes the VM lifecycle processes from request through provision, operation and resource recycling.

VRM’s self-service provisioning manages communication between end users and IT through a common portal, automating the request and approval processes. It also automates the provisioning of VMs, including the selection of physical resources, which leverages policy driven compliance to limit resource access and assures VMs are built to standards. VRM enables authorized VM provisioning in minutes, usually without manual intervention.

VRM provides a dashboard control panel and operational reports to facilitate the management of the virtual infrastructure. It provides a graphical representation of operational status and resource utilization. Reports show configuration details, audits of end user activity and workflows.

Since most environments include more than a single platform, VRM works across virtualization technologies from Citrix/Xen, Microsoft, Sun and VMware. Built around Microsoft’s .NET 3.0 and Windows Workflow Foundation process automation, it’s easier for partners and users to customize workflows with VRM, as compared with a proprietary dynamic operations engine. VRM manages virtual desktop implementations, as well as servers and can scale to 1000s of machines.

Storage Switzerland’s Take

Lifecycle management of a resource is not a new concept. It originated years ago as a strategy to optimize storage and align data value with storage costs. But resource optimization is somewhat of an ideal, rather than an imperative. As long as the cost to waste a resource is below the cost of implementing efficiency, we’ll continue to see ‘suboptimization’. This was, and  certainly is the case in storage, as Moore’s Law continues to drive down the per GB cost of disk and companies keep buying more.

VMs are a little different than storage capacity. These aren’t unused files, they’re server instances, that each include a resource commitment as well. Until those VMs are decommissioned, they’re consuming storage, memory, CPU and bandwidth on the network - not to mention admin time. And, unlike disk storage, Moore’s Law isn’t making these resources cheaper every year. VM lifecycle management is a process created to address the problem of VM sprawl by controlling the creation, management decommissioning of virtual machines, and the reclamation of their associated resources. 

Proactive management of the virtual compute environment is an imperative and tools like Virtual Resource Manager are needed, as traditional monitoring tools, which report after the fact, will be ineffective at curbing VM sprawl. Left to traditional methods and control procedures, VM growth can take a similar trajectory as that of data storage. But, without the ongoing cost reductions that disk storage enjoys, this growth isn’t sustainable. VM lifecycle management is required if organizations expect to stay on top of their IT environments as virtualization continues to expand.

Eric Slack, Senior Analyst