The KB article that I reported on was incorrect. I’ve since posted a correction that shows you what the KB article should have said.
Changes and improvements in Windows Hyper-V Server 2012 no longer support pass-through disks if Live Migration is to be used.
I, for one, applaud this. I’ve been quite vocal about disliking passthrough disks. If you need the flexibility and the fastest storage performance then use fixed virtual hard disks. A fixed virtual hard disk is almost the same speed as the underlying LUN and you’re probably getting lots of IOPS by stacking disks in a LUN or disk group. Half the Internet (if not more) runs like this, so stop thinking that your mission critical app for 100 users needs to be special – especially if you’re guessing … and I know you haven’t done an assessment. You know who you are.
Thx for the Post. I dont like pass-through disks too.
Do you have some experience with Hyper-V 3 Npiv?
I maybe want use that for some Guests.
No.
This makes me nervous when dealing with greater than 1TB disks. We end up with giant files (VHDX) that we have to manage (backups, Hyper-V Replica, migrations) and hope that they do not become corrupt in some way. Corruption on a single file and a loss of 1TB of data?
Backup.
not cool…
We use this to allow san snapshots to be application aware.
Seems like we will have to use iscsi lun connected within the vm to be supported.
I suggest you learn how Hyper-V backups work. There is a Hyper-V VSS writer for application consistent backup of VMs.
Aidan, I completely agree. Pass through disks offer little benefit at tremendous cost. The one time I used a pass through disk I was amazed at all the flexibility I lost and I saw no performance gain.
JamesNT
As does Microsoft storage PM, Jose Barreto, according to a tweet by him today.
I use pastrough disks for data on virtual node of SQL cluster. One node is physical and one is virtual (passive). I will need shutdown virtual node on maintenance now.
You should have listened to those of us who’ve been telling you not to use passthrough disks.
I agree that NOW ( not 2 months ago, NOW) that we have DPM2012 SP1 + all the fixes and differential backup of hyperV ( i know they’re NOT differential but 900% of improvement made look that they’re differential in comparison ) VM MAYBE passthru disk are a thing of the past.
But give the fact that DPM201X is so efficient at backup file server ( and there are huge file server around the world ) SQL Server Exchange Server and Sharepoint, and it is not possible ( afaik ) backup only some vhd in a VM we have to resort to BRM of some machine even if virtual machine this is not optimal ( and you have to try do a secondary DPM backup of BMR to see why ).
For me, simply NOT ENOUGH to remove support of this feature.
Plus you can resize a VHD online? not that i’m aware, so even fixed WHD have at least a serious pitfall.
But simply they cannot claim a thing like this after month from RTM.
If they plan someting like this, they must declare at least 2 major releases in advance.
The fact is that they have problems to do it and try to cover it with some [replace with tour own curse].
Interesting… DPM 2012 best practices state you have to use a passthrough disk and not a VHD or VHDX. I’ve been using this configuration for my DPM VM since October 2012 and havent had an issue with live migrations. I’ve had other issues with SCVMM 2012 recognizing a VM with a GPT passthrough disk, but never an issue with live migration.