VMware Blog: Scaling real-life Web server workloads – VMware
In this scalability test they compared the performance aspects (such as latency, throughput and CPU resource utilization) of real-life web server workloads in a native environment and a virtualized data center environment.
KVM VDI Offering – the Register
Qumranet, the well-moneyed entity behind KVM, is asking folks to help it test out some new desktop virtualization code that sends applications out from headquarters to remote offices.
Back in April, Qumranet birthed SolidICE, which is the the company’s take on desktop virtualization. Like rival software, SolidICE ships desktop software from a data center to a user’s thin client or PC. The big buzz around such a code is that it makes managing, upgrading and securing lots of desktops easier. The downside, however, has been that applications funneled out from the data centers tend to lag once they arrive on the desktops, much to users’ chagrin.
VMware Release New Virtual Disk Toolkit – run-virtual.com
VMware has released this week a new toolkit for virtual disk management that is looking very promising. The VDDK provides easy access to virtual disk storage. This enables a wide range of solutions including:
Microsoft HyperV: Microsoft Assessment and Planning Toolkit 3.1 for
The Microsoft Assessment and Planning Toolkit 3.1 (or MAP) is a network-wide agent-less tool that can help you quickly find out where your desktops and servers are as well as auto-generate upgrade recommendations for multiple products and technologies including server, desktop and virtualization migration scenarios covering:
HyperV benchmark FUD:
Well, I suspected the FUD machine has become active. Chris Wolf on his virtualization blog debunked a Qlogic/Microsoft bench mark that showed high IOPS for storage access. Trouble is to get these high values they had to use solid-state disks! Yeah, like we all have those in the average datacenter
Similiarly at the MS virtualization blog they repeated an old-chestnut claim about their “64-bit” hypervisor
“The Hyper-V model allows for much better I/O performance of virtual machines. This is mainly due to the new 64-bit hypervisor layer underneath everything – including the host operating system even, and the new high-speed VMBus “synthetic” drivers that run in the virtual machines”
Might not know that many hypervisors (including VMware’s VMkernel) are based on 32-kernel – and it makes NO DIFFERENCE TO PERFROMANCE. Why, because the virtual machine makes native calls of the underlying CPU virtually bypassing the hypervisor altogether. It’s called “direct execute” in VMware documentation – and VMware were one of the first to support 64-bit where it counts – within the guest operating system. 64-bit Windows never had the popularity of 32-bit, because it is a pain to get all the device drivers for a physical world. Ironically, the pioneers of virtualization (AKA VMware) have more easily adopted the 64-bit guest operating systems. It’s much easier to setup 64-bit Windows in a VM, than in a PM.
Does HyperV kill Citrix Xen, leaving KVM in the driving seat?
David Marshall of vmblog.com does a good summary of Brian Madden’s claim that Microsoft’s Hyper-V will “kill” Citrix Xen. Those after a premium product will choose VMware, whereas cost sensitive deployments will choose Microsoft . Madden then predicts the open-source folks will focus on KVM as an alternative to Xen. Now, where’s my crystal ball. If I could see the future like this – I would take an accumulator on the World Series, Championship League and the FA Cup!
Installing ESX 3.5 inside a VMware Workstation Virtual Machine
Title says it all. Video on the vmblog.com
This was first published in July 2008