Hi,
I am just taking my first foray into Virtualisation, and have successfully virtualised a phyiscal Terminal Services Server using ESXi 3.5. It's running a treat and I am really pleased with it. But before I start virtualising more machines, I need to be confident of my backup/restore procedures.
It seems the most simple way would be to store the VMs on an NFS store, then re-import them onto a newly provisioned host, should my current ESXi host fail. I can subsequently create backups of the VMs that are stored on the NFS also.
Does this make sense? Will my terminal services server get noticeable slower if I move the VM to the NFS data store?
Welcome to the Forums - It will depend on the NFS datastore you will be using - for example there are many VMware customers using NetApp NFS as their shared
storage with no performance issues
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
We are using a Volume that exists on our network that is shared out from an Apple Xserve (OSX 10.5) using NFS. So I guess I may not find the answer unless someone else is in a similar scenario?
If I attempt it myself, I will post back here with results.
I haven't seen any performance issues with VMs in NFS, and our cluster has over 400 VMs in different filers. Just make sure you use a good filer and at the very least a 1Gbps connection.
Check this performance comparison:
http://media.netapp.com/documents/tr-3808.pdf
If this post has been helpful/solved your issue, please mark the thread and award points as you see fit. Thanks!
Thanks for the pointers piaroa;
I guess I am gonna have to suck it and see, I suspect that the VM that I am considering moving will take some hit but not too much as I don't think it has a lot of hard disk I/O.
Might make the move tonight.
BTW forgive my newbie ignorance but can you explain "make sure you use a good filer"
I mean, don't throw a high number of running VMs into a home-office filer (Iomega, netgear, Qnap, etc), running on SATA 7.2kRPM disks.
Maybe I should have said an enterprise grade filer, running at least 10.000K rpm disks for IO intensive VMs.
If this post has been helpful/solved your issue, please mark the thread and award points as you see fit. Thanks!