Hyper-V Program Manager
I had a great time today at TechEd talking about Dynamic Memory in Hyper-V R2 SP1. I am going to do some detailed posts about it in the coming days – but for now I am just posting the presentation files that I used today (click on the title of this post to get the files).
I enjoyed your slides, thanks!
I wonder if, in your coming detailed posts you could answer this question for me?
I have a Server 2008R2 Cluster (3 nodes) running 9 highly available hyper-v VMs, they all have 24GB or memory, at the moment I need to leave at least 10 GB of memory free on each of the nodes so that if one fails the other two can start the failed-over VMs. With dynamic memory, could I utilise this idle memory and re-assign it to the migrated VMs in a failover situation?
Seems like such a waste having all that memory sitting idle.
Thanks for your help.
Why, oh why can't you guys put notes in your slide decks? Now I'm going to have to wait a few weeks to get all of the details because my employer is too cheap to send us to TechEd.
Thanks Ben for the great presentation. Really enjoyed it. Especially that you looked like an (uncomplicated, friendly) human being, compared to the keynote...
The only question that remains for me:
If I run a server application that uses all available memory for caching (SQL Server, Forefront TMG, ...) how will this work with Dynamic Memory?
Windows tries to use all memory for caching and Superfetch, but as you explained in the talk Hyper-V just looks at the right performance counters to determine the actual memory need. As far as I can tell this is not possible for applications.
The video is amazing, but at the end, you should repeat the question made by the audience, because we just cant understand what they are saying. They dont have mic... but anyway, it's a long video, but I leaned a lot. Thank you!!!!