I’ve been watching an interesting “discussion” take place over the past few weeks via blogs, and honestly, I think it’s pretty funny. At the same time it is also pretty sad. More than a few years ago, Microsoft took the approach of trying not to criticize a competitive product, but to try and highlight our products’ values instead. I’m not saying we have always been successful, sometimes we have made poor decisions in this area, but we at least try to provide the data that influenced our position. If we are list pricing differences, we can reproduce the details of how we came to the pricing information. If we show performance statistics, we at least try and describe our test scenarios so others can reproduce it as well.
As a whole, I think we’re making progress in this area: That brings me to the following blog / You Tube posting / press article. A while back a video of a Hyper-V server blue screening was posted on You Tube. You know You Tube, you can find a video of almost anything. Well this one got under the skin of our Hyper-V product managers and they asked for clarification. Hyper-V has been incredibly robust and a blue screen of this magnitude would be considered a show stopper.
Here’s Jeff’s latest posting regarding this situation
Here’s another one of Jeff’s postings that provide additional background on the situation
If you haven’t read enough, another nice note from Jeff pointing out some of the additional background around performance and uptime
Here’s the latest press article on this topic. Hmmm… While I usually try hard to stay out of the politics of things like this, this one just won’t go away. Please let me know what you think.
Here’s another blog for additional background. As I said earlier, we don’t take the “high road” as often as I’d prefer, but we at least provide the background for how we assembled our data. This blog is from Keith Ward, check it out.
Slinging the YouTube Mud
Here’s an excerpt from Keith’s blog that includes the link to the You Tube video.
Now, the YouTube is on the other foot. A video has been posted of VMs in Hyper-V repeatedly crashing, followed by a bluescreen of the host itself. It's pretty scary stuff if you're using, or considering, Hyper-V.
So where does this stand? We still don’t have enough detail about the VMWare tests to understand how this happened. We’ve asked VMWare to provide data so we can dig into this. If we have a problem, we’d like to fix it. We have not been able to reproduce the crash on our own. We’ve asked for the dump data, we’ve even asked for a detailed explanation of the configuration. At what point do we just need to say that the video is is a fabrication? Anyone, with the appropriate level of access to a software installation can make that piece of software crash. With enough time and energy man can destroy anything, but aren’t we all in the business of trying to make things better? How many customers have their Hyper-V machines blue screen? For that matter, how many customers have their VMWare servers crash? I’d wager to say, virtually NONE! Why? Because our job as professionals is to put our customers in a position to succeed, not fail. Let’s focus on some success stories ok?
This leads me to another issue that has rubbed me the wrong way. Wouldn’t you like to see a comparison of Hyper-V and VMWare head to head? To me, a head to head comparison is a success story waiting to happen. If you’ve read all of the posts above, you’ve seen the comment that VMWare’s EULA does not allow public benchmarking without VMWare’s approval.
What if both companies assemble their environments on like hardware and go head to head? Similar to Presidential debates? I’d love to see it, but I suspect the political wrangling between the two teams to try and ensure the environment is “fair” for everyone will prevent this from happening. I believe any fair test like this would be a win-win for both companies… Why? I’m not naive enough to think that Hyper-V will win every test, there are places that VMWare does excel and probably will keep their advantage, but I think a head to head would highlight where each solution excels. To me, this is the success story in the making. I honestly believe that there are scenarios where VMWare is the better solution, but I also believe that there are plenty of scenarios where Hyper-V will also excel.
All I can say to my partners and customers is that you can always test the performance of VMWare and Hyper-V. You just cannot disclose the results outside of your company; but make up your own mind. I ask, that for the benefit of your company, you do your “due diligence” to configure each environment to the best of your ability. You won’t be doing anyone any good if you build a test scenario that inherently disadvantages one of the competitors. Use your hardware, use your scenarios, and make up your mind.
One last thing: When your test is complete, look at the cost / performance numbers. I suspect that VMWare will win in some categories and Hyper-V will succeed in others, but at the end of the day, look at the cost of the performance you achieved. Look at the overall cost of the VMWare solution and compare it to the overall cost of your Hyper-V solution. Which one costs more to deploy? Which one costs more for end to end management? And which one costs more for the lifecycle of the solution?
That’s enough for now, please let me know your thoughts on this whole “Kerfuffle”…
Until next time,