So some follow up on my research. I wish there was a guide compiled with all this stuff, for us newbies it is hard to consolidate it. It does appear like it is reasonable to put your guest OS paging file on a separate volume.
1. add disk – set to persistant
2. boot VM – format new disk and set pagefile
3. boot VM again – now Windows starts to use the pagefile – do nothing and shut down again
4. now edit vmdk properties and set pagefile.vmdk to nonpesistant
I really can’t find any offical VMware documentation thought that this is a good thing?? Although this article alludes to doing it. It seems good especially since it won’t be part of the snapshot anymore.
Using separate physical and virtual hard disks Install the host operating system onto a separate hard disk than the virtual machines. Also store the paging file or swap partition on a different drive than the host operating system.
With a little more poking around it looks like this is ‘Best Practice’ for NetApp so if its good enough for them it should be good enough for us right!
As NetApp best practices for VMware recommends, transient and temporary data such as the guest
operating system swap file, temp files and pagefiles, should be moved to a separate virtual disk on a
different datastore as snapshots of this type of data can consume a large amount of storage in a very short
period of time due to the high rate of change.
So after reading that I’m going to put the temp files on that volume as well. I’m also considering putting the windows log files that separate partition as well, anyone do this?
So then it looks like you should also get your ESX swap files onto their own partition as well.
For some reason it seems to be hard to find this as a best practice as well.
This lays it out pretty well
Although I can’t find it mentioned it seems (obvious?) that you want to but the ESX guest swap files onto a shared volume so that you can still vmotion them!
I’m not sure what I think of the complexity of all of this, it starts to make a VM very complex especially if you have any iSCSI connected volumes from within your VM as well!
Great write-up. Keep us all posted what works well, what doesn’t work well, and what doesn’t help or work at all. I too am unsure what to think about the added complexity, but if the intent is to make replication efficient, then these things need to be considered.