Virtual servers cause bottlenecks

June 20, 2009 – If you’ve moved beyond test/dev and into full production with your virtual server deployment, you’re well aware of the fact that as ‘VM sprawl’ creeps in performance goes down. And you’ve also probably figured out that in the majority of cases, SANs are the culprit.

In fact, in a presentation at the VMworld Europe show in Cannes earlier this year (no, I didn’t go), Scott Drummonds, VMware’s group manager, technical marketing, said that 90% of the performance problems in VMware environments were SAN-related.

In researching an InfoStor Special Report on SAN management in virtual server environments, I checked with some industry analysts and end users and that percentage may be closer to 70%, but nevertheless it’s clear that in order to maximize performance of virtual server farms you have to maximize storage performance.

But first, you have to monitor and analyze performance across your entire virtual infrastructure, particularly the SAN fabric. In researching the Special Report I came across some tools that you should look into before you actually try to tune your environment, because “you can’t optimize what you can’t measure” as the folks at Virtual Instruments say.

Those tools include Akorri’s BalancePoint, NetApp’s SANscreen VM Insight, and Virtual Instruments’ VirtualWisdom (which is based on the company’s NetWisdom and is currently in beta, with general availability slated for the fourth quarter).

The Taneja Group research and consulting firm refers to these products, and others, under the umbrella term ‘virtual infrastructure optimization, or VIO.’ These cross-domain monitoring and analysis tools provide very deep visibility into the virtual infrastructure, with an emphasis on performance analysis and optimization recommendations.

I spoke to SAN administrators that were using some of these tools and I’d highly recommend at least taking them for a test drive if you’re having any performance problems in your virtual server implementation.

Also check out a recent Webcast, archived at infostor.com, titled Eliminating the I/O bottleneck that limits scalability in today’s virtual servers. The main presenter was Greg Scherer, Neterion’s CTO, and although it does have some Neterion-specific content it’s an excellent look at I/O issues that often crop up in virtual server environments.

That Webcast was the first in a four-part series on Virtual Servers and the Impact on Storage, brought to you by InfoStor and our partner, Virtual Strategy Magazine.

The next Webcast in this series is coming up fast: On Thursday, June 25, 10:00a.m. PDT, 1:00p.m. EDT, we’ll host Virtual Servers: Six Ways to Reduce Your Backup Window. The key presenters for this Webcast will be executives from Vizioncore. To register, click here.

I’ll keep you posted on future Webcasts in this series.

Labels:

posted by: Dave Simpson

Dave Simpson, Editor-in-Chief
by Dave Simpson
Editor-in-Chief

Dave Simpson has been the Editor-in-Chief of InfoStor since its inception in 1997. He previously held editorial positions at publications such as Datamation, Systems Integration, and Digital News and Review. He can be contacted at dsimpson@quinstreet.com

Previous Posts

Archives