Q: If several client networks are virtualized through one card, what does this mean for low-latency applications?
Meet the expert
Martin MacLeod is an IT consultant based in London with several years' experience deploying and supporting blade servers in volume. He also blogs at bladewatch.com, where he tracks news on blade servers, grids and virtualization. Listen to the rest of Martin's answers on blade server virtualization by downloading our blade server virtualization podcast.
You're going to have to be a lot more careful about where your applications are running, how you're going to split up the network between the virtual machines, and just make sure you're getting the same level of performance from a network standpoint or from a service standpoint through the network ports. This is because a lot of times you're teaming the network part, which means that you might want several network ports on a blade, which is where you'll get different vendors offering different network port daughter cards. For low-latency applications, you just need to make sure that you optimize the [VMware] ESX configuration to see what priorities you can set at the ESX level and at the network level to maintain service.
TechTarget provides enterprise IT professionals with the information they need to perform their jobs - from developing strategy, to making cost-effective IT purchase decisions and managing their organizations' IT projects - with its network of technology-specific Web sites, events and magazines.
