For years, I’ve been meaning to get some business cards sorted for these VMworld Events, I always feel a bit of muppet when someone hands me there’s and scribble my name down on a napkin! Anyone our friends at vInternals have beaten me to the coolest business cards… just check out these images!
Now, that’s a man after my own heart! Keep it Up! 
[This was a meant to be post, not a page - opps!]
Well, I’m down at the bookshop with my book. Got an hour or so to go. The book has been discounted now at €25. So come on down and grab a signed copy whilst stocks last!
This session was about VC Heartbeat Service. It’s not free (its actually quite expensive!) and I believe its a licensed version of NeverFail for VirtualCenter. It looks like you cannot VMware FT vCenter (don’t quote me on that – this still bleeding edge stuff which hardly anyone has tried!) which would be the natural thing to do. So basically, VMware create single-point of failure for MANAGEMENT (I know VC is not single point of failure for VMs)… upon which ALL the high-level management layers (SRM, LCM, LM, View, etc) depend – so not only I have to pay for vCenter, I pay again if it fails… Yikes!
I’ve got customers who have had VC up for 18 months without a problem (they clearly do no upgrades then, Mike?) so you gotta wonder whether its worth the spend…
In this session we were bored rigid for a bit about what ESX currently does – things picked up a gear when the guys talked about future developments… Yes, the Service Console will have a virtual disk… that’s if you still using ESX Classic, as opposed to ESXi
My book is in the book store on SRM – Tommorow I will be doing a signing of the book from 12noon to about 2pm…
This session discuss the affect that VMware FT has on performance. 3 main things – firstly, you can only current protect a 1xVCPU VM with FT. Given that one way to improve performance is add more vCPU that needs to be kept in consideration. Currently CPUs don’t really have a hook to allow VMware to track the memory changes of two vCPUs – whereas the “lockstep” attribute on most modern CPUs do allow this for single vCPUs. 2ndly, latency is the key – the gap between the Primary VM and Secondary VM because the Primary VM cannot massively go quicker than its secondary. Whilst latency is small for outbound traffic (because that is not replayed in the secondary) high amounts of inbound network traffic can have significant impactly. 3rdly, very I/O bound VMs for storage may need switching to secondary method which in test didn’t out perform the default method – this secondary method is labelled as “shared read mode”…
This is just small except of a demo of a VM running on a Nokia Mobile phone…
In this keynote session Steve Herrod – outlined the future of VMware virtualization. The coolest part being the demo of a virtual machine running on a mobile phone – shame it was Windows CE! Hee-hee
VMworld Europe - 2009: Episode8: Future of Virtualization Keynote Play Now | Play in Popup | DownloadHere’s my report for the hotel room of the 1st day. In the evening I was out with Hyper9 who put food in my tummy. Got chatting to Gabe’s babe who’s antropologist – who is “studying” VMworld like we are some lost tribe in the jungle. You know what there maybe some truth in that!
Here’s a video of me and Mr Eric Sloof falling asleep within about 30 seconds of the SAP part of yesterdays keynote…. People were leaving in droves!
http://blog.virtualarchitect.nl/2009/02/vmworld-europe-2009-tuesday-keynote-bonus/
If you want to add Mike Laverick on LinkedIn, click on this button:
RSS
iTunes