ZFS dedupe!
Well, its been a few months since Sun announced ZFS getting dedupe functionality. Now I just need to get employed again, so I have the cash to build the home storage server I've been meaning to setup for some time now. Add another processor/memory heavy machine to run ESXi, and I'll have a nice little virtual machine lab setup so its easier to study for my certifications by having every OS handy to experiment and research with.
I occasionally have too much free time…
In the effort to reduce the accumulated clutter in my bedroom/office, I took a few minutes to decide if some things could be re-purposed elsewhere in the house. A spare thin client terminal, a LCD monitor and a keyboard/trackball combo were sitting near each other, and a thought occurred to me. I already have VMWare Server running on my desktop, hosting some VM's (virtual machines) for various certification and testing labs, so I setup another VM desktop, and carried the equipment downstairs. I hooked everything up and plugged the thin client into the home theater's data switch, and voila, a terminal to surf for recipes in the kitchen! Taking a step back, however it looked kinda ugly with the network cable running around the wall, so its temporarily removed until I can run a cable inside the common wall to a faceplate to tidy it up.
vSphere 4′s improved vCPU scheduling: performance improves across the board
VMWare's whitepaper posted in August, The CPU Scheduler in VMware ESX 4, shows how they have significantly improved the performance of their vCPU (virtual CPU core) scheduling at both light loads and heavy loads (performance stays about the same in between though). Light load performance improves through better scheduling on idle pCPUs (physical CPU core) with avoidance of unnecessary pCPU migrations, and heavy load performance improves thru less scheduling overhead, finer grained vCPU scheduling, along with better decision making for vCPU migrations between pCPUs.
In ESX 4, many improvements have been introduced in CPU scheduler. This includes further relaxed co-scheduling, lower lock-contention, and multi-core aware load balancing. Co-scheduling overhead has been further reduced by the accurate measurement of the co-scheduling skew, and by allowing more scheduling choices. Lower lock-contention is achieved by replacing scheduler cell-lock with finer-grained locks. By eliminating the scheduler-cell, a virtual machine can get higher aggregated cache capacity and memory bandwidth. Lastly, multi-core aware load balancing achieves high CPU utilization while minimizing the cost of migrations.
Excellent & Free VSphere tutorial videos!
Christopher pointed out these great videos by Mike Laverick to supplement his forthcoming book. I know what I'm doing for the next few hours! They're in Flash video format at 1024x768, so most netbook users won't be able to see all the action, and there's no transport controls, so make sure you have enough time to finish the clip when you start it (the clips vary in length from 10 to 30 minutes), or you'll get to "review" the content when you come back to finish it.