Why is cost management essential in hybrid IT environments?
Cost Management is essential in Hybrid IT (virtualized and cloud) environments for two reasons:
- Virtualized environments become less cost effective with use over time, and
- VM costs varies significantly based capital and location
1. Virtualized environments are not cost effective:
While virtualization has brought flexibility and ease of operation to enterprises, one of the great promises of virtualization – significant cost savings through improved infrastructure utilization – has not been delivered.
This is because as virtualized workloads continue to run over a period of time utilization degrades due to several reasons:
– Over-provisioned VMs: To prevent performance issues, many workloads are over-provisioned with virtual resources that are unused, and therefore leads to wasted costs
– VM sprawl: More VMs than are necessary are provisioned to workloads, once again, to prevent performance issues, or because the utilization characteristics of a workload are not well understood.
– Poorly configured VMs: Often workloads use one resource more heavily than the other. The workload may be memory intensive or CPU intensive, but the VM supporting it includes a healthy allocation of unnecessary CPU/Memory/Network/Storage.
Many studies have shown that virtualized enterprises become 30-60% under-utilized over a period of time.
2. VM cost varies significantly based capital and location:
All VMs are not created equal. A VMware VM running on UCS blades in a Windows host in a data center in North America costs much more than say a Xen VM running in a Linux host on Dell blades in a data center in Bangalore.
Most enterprises do not have visibility into the actual cost of operating a VM. This visibility is crucial to allow for informed prioritization of allocating workloads to VMs as well as visibility into actual cost savings from optimization.
So, enterprises need to perform cost management on a daily, weekly or monthly basis to:
Proactively prevent significant cost drain from badly configured VMs, and
Understand the $$ value of various VMs to make informed decisions around allocation and optimization.
Who should read this white paper: Any IT manager who manages physical and virtual workloads whether they are running in-premise or in-cloud and would like to measure the cost of running each workload for the purposes of:
– optimizing the infrastructure to save costs
– planning for migrations of workloads to other platforms/sites/clouds