Poor Hyper-v performance on guest machines

Hi Experts.

I'm having performance issues with our Hyper-v server. I'm not very experienced with HYPER-V so I was wondering if you guys could help me out on this one.



Windows Server 2008 R2 SP1 64Bit (full install)
2x Intel Xeon X5690 @ 3.47GHz with 12 cores each
32GB of Memory
5x 1TB HDDs, split in two mirrors and one hot-spare
C: is one mirror, D: is another mirror
2x NICs: Intel NIC and Intel NIC#2

Intel NIC: "Vantage Ethernet Connection" -  static, G, DNS
Intel NIC#2: "Local Area Connection 2 - Virtual Network" - no gateway, DNS

OS and roles are installed on C:

Roles installed:

File Services

Hyper V guests:

Guest #1:

Windows Small Business Server 2011
Domain controller, Exchange, File Server
Dynamic Memory (minimum 8GB maximum 24GB)
Physicall files on C:
Uses Local Area Connection 2 - Virtual Network, IP address / 24

Guest #2:

Windows Server 2003
Line of Business application running Database
Static memory (4GB)
Physical files on D:
Uses Vantage Ethernet Connection, IP Address / 24

Here are the issues they are describing:

1. When you perform any operation on Outlook there is approximately a 40 second delay. When they click to send the Email they will have to wait 10 to 40 seconds for the window to close
2. The line of business application my client uses is very slow to go from section to section, which connects to the Server 2003
3. Opening cad files from the network and saving them is very very slow

SBSSRV cpu is running at a good 5%, and memory at 87%
HYPER-V HOST is running at a good 5% and Physical Memory 48%

I have attached a sreenshot of the Performon results for both HYPER-V HOST and SBSSRV

Also attached a resource monitor screenshot also for both HYPER-V HOST and SBSSRV

I think the issue is either Disk or Network, but im unsure of what to do next to fix the issue.

Thanks.   Hyper V Performance monitor Hyper V Resource Monitor Hyper V Task Manager NICs SBS  Performance monitor SBS Resource Monitor SBS Task Manager
Craig RutherfordAsked:
Who is Participating?
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Snapshots are not supported running in Production. and you should not really run on a Snapshot for more than two days, or you are going to get in a real mess, with lack of storage space, or not being able to committ the snapshots, I would stay clear of using Snapshots.

and when in Snapshot mode, (AVHD) mode, ALL VMs running a Snapshoty will run poorly.

Anti-Virus, On-Access scanning is not good either.
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
If those are 1TB 7,200rpm SATA drives, the bottleneck is likely to be with the storage, which is causing you performance issues with your VMs, which is then hitting the CPU.

2 x RAID 1 Mirrors is not going to give top performance.

You would be better with 4 x DISKs in RAID 10, with a hot spare.

More spindles, more disks = more performance.

But SATA 7,200rpm disks are troublesome, because they cannot give the IOPS needed, upgrade to SAS, or purchase more SATA disks.
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
It's a shame because you've got a top spec machine, 32GB, Dual Xeons, 12 Core each.

It's unlikely to be network, it takes alot to saturate a 1Gb network.

What's the Storage system SATA, 7,200rpm?
Improve Your Query Performance Tuning

In this FREE six-day email course, you'll learn from Janis Griffin, Database Performance Evangelist. She'll teach 12 steps that you can use to optimize your queries as much as possible and see measurable results in your work. Get started today!

Craig RutherfordAuthor Commented:
The thing is that I have a very similar configuration in another client of many, with a HyperV host and 2x Guests, one running SBS 2011 and one running Terminal Server (25+ concurrent users). Their hardware is much smaller than the one described above, and they don't experience this bottleneck, eventhough they also run 2x Mirrorrs
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
SATA 7,200rpm disks?
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Does your other client run a database?
Craig RutherfordAuthor Commented:
Hitachi HUA72201 SATAII 7200 3.0 gbps
Craig RutherfordAuthor Commented:
no, no database on the other one
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Yes, it's possible the database, which is cuasing the issue on SATA drives. Sorry, but that's the bottleneck, good server, let down by poor storage.
Craig RutherfordAuthor Commented:
I think I just found what the issue was. We have an Endpoint Security Antivirus from our Monitoring Agent (n-able), that uses the panda enginem installed on both SBS & Database server. I have just uninstalled it completely and the HDD usage dropped by nearly half. I find it strange it didn't come up on the resource monitor, but maybe I just wasn't looking long enough.

I will be performing some more tests and will get client to test the performance as well. Will keep you updated.

We have about 5 snapshots in one of the servers as well and im thinking of removing them to see if it helps too.... but I have to have a good read about it, I'm not sure if its as easy as just deleting them, of if deleting may revert the server back to the initial state.

Will keep you posted.

Thanks for the help mate.

Craig RutherfordAuthor Commented:
Hey guys, I'll post the results for this tonight, it seems to be running ok now.

Speak to you soon
Craig RutherfordAuthor Commented:
Hey mate, forgot about replying to this.

The problem really was the AV as I said... thanks for the assistance and pointing the right directions
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.