Fine-tune your virtualization performance management skills
A comprehensive collection of articles, videos and more, hand-picked by our editors
Slow virtual machine performance is a problem to which every virtualization administrator can relate. Slow performance is often a result of resource contention or shortages caused by unforeseen demands. The slower access speed of spinning disks make them a potential bottleneck and failing to allocate enough storage to a VM can cause serious problems. Luckily, administrators have several strategies for improving VM performance trouble caused by disk access. Here are five quick links to help you troubleshoot and improve VM performance related to storage I/O.
How disk latency and RAID groups affect VM performance
It's easy to forget that both disk latency and RAID level can affect VM performance. The very fact that a disk relies on spinning platters often makes it the limiting factor when it comes to performance, but spreading data access across more than one disk can reduce seek time. However, keep in mind that some RAID levels will actually introduce more latency.
How storage location affects VM performance
Storing VM data on the local server can have the obvious performance advantage of avoiding network bottlenecks and latency. But, those small advantages also come with big drawbacks, including limited VM mobility and difficult management. In most cases, it makes sense to store VM data on a storage area network.
Improving performance through proper disk selection
Getting good performance starts with having the right hardware and correctly allocating that hardware to VMs. Follow these rules of thumb for allocating storage and you can avoid potential problems.
LUN configuration best practices to boost VM performance
With proper LUN (logical unit number) configuration, you can improve VM performance. Learn how to choose hardware, RAID levels, disk types and VM storage technologies for your LUN configuration.
How many VMs should you put on a LUN?
The number of VMs you can put on a LUN and still expect optimal performance varies based on the size of your infrastructure and whether the VMs are for development or production. If you assign more VMs to a LUN than your infrastructure can handle, you can quickly saturate disk I/O.