Skip Links

Be careful using VMotion over the WAN

Wide Area Networking Alert By Jim Metzler and Steve Taylor, Network World
May 24, 2010 12:03 AM ET
Jim Metzler
Sign up for this newsletter now!

Insightful analysis by consultants Steve Taylor and Jim Metzler, plus links to the latest WAN news headlines

There is no doubt that server virtualization has been one of the hottest topics in IT over the last few years. In addition to the CAPEX and OPEX savings, one of the compelling reasons why IT organizations implement server virtualization is the ability to move virtual machines (VM) between physical servers. As we will explain in this newsletter, moving VMs between physical machines may not work if the VMs have to traverse the WAN.

VDI cheat sheet

As noted, one of the real advantages of server virtualization is that it enables a production VM to be transferred to a different physical server, either to a server within the same data center or to a server in a different data center, without service interruption. This capability enables workload management and optimization across an IT organization's virtualized data center(s). This capability also helps to streamline the provisioning of new applications, improve backup and restoration operations and enable zero-downtime maintenance. VMware refers to this capability as VMotion. Our research indicates that over 40% of the IT organizations that have implemented VMotion use the capability one or more times a week.

However, according to Cisco, VMware and NetApp if you want to move VMs over the WAN you need at least 622Mbps of bandwidth.

While that is a lot of capacity, that is not our primary concern relative to transmitting VMs over a WAN. Our primary concern is that those vendors also state that the maximum round-trip latency between the source and destination VMware ESX servers cannot exceed 5 milliseconds.

The speed of light in a combination of copper and fiber is roughly 120,000 miles per second. In 5 ms, light can travel about 600 miles. Since the 5 ms is round trip delay that means that the data centers can be at most 300 miles apart. That 300 mile figure assumes that the WAN link is a perfectly straight line between the source and destination ESX servers and that the data that is being transmitted does not spend any time at all in a queue in a router or other device. Both of those assumptions are unlikely to be the case.

We asked Cisco about this issue and their response was that the 5 ms constraint was a result of both the TCP buffers and the synchronous vs. asynchronous mode on the storage devices. Cisco went on to say that work is underway to address those issues. We believe that Cisco and others are working hard to resolve this issue in part because it is in their best interest to do so. However, in the meantime, if you want to run VMotion over the WAN be careful, be very, very careful.

Read more about lans & wans in Network World's LANs & WANs section.

Steve Taylor is president of Distributed Networking Associates and publisher/editor-in-chief of Webtorials. Jim Metzler is vice president of Ashton, Metzler & Associates.

Our Commenting Policies
Latest News
rssRss Feed
View more Latest News