Dexter Filmore wrote:
Currently taking part in an MCSE certification training, (yes I know. but LPI
C2 doesn't get me a job here. Nothing I can do about it) and situation is as
follows:
At the facility I got an intel E2140 (that's the crippled Core2Duo with 1MB L2
at 1.60 GHz) with 2GB RAM running Win2003 server.
On that VMware WS6, on that a range of VMs sitting on a Samsung sATA, NTFS,
nothing fancy.
VMs are: Win2003, XP Pro, Vista, BackTrack3/HDinst (had to make myself at home
*somehow* :) )
This machine restores 2k3, xp and BT3 *simultaneosly* in 15 seconds to the
point when they react to mouse input, another 60 seconds till they are fully
loaded.
My box at home:
AMD Athlon 64 5600+ 2x2.90GHz, 512k L2, 4gigs of RAM, same HDD, /dev/shm
mounted,, Kubuntu 8.04 64b and debian etch, VMware WS6, VMs on an xfs
partition.
Restoring the 2003 Server alone takes roughly 3 minutes, and it won't react
smoothly before that point. I notice a lot of harddisk activity. Both
partition and guest OS vmdks are defragmented.
Starting 2 VMs simultaneously is a total mess.
(It used to work a lot better a year ago on Slackware 11 but I don't have that
installed anymore, can't compare.)
Anyone got pointers on how to spot the bottleneck?
I'll try moving the server VM to another disk tonorrow, I somehow suspect xfs,
then again xfs was made for handling large file afair.
How big are the VM's in terms of emulated RAM?
A 1GB virtual machine should take at least 20 seconds *just to read the
RAM from disk*.
I would check the VMWare forums. This question has likely come up there.
I suspect that they heavily optimize the Windows version and just leave
the Linux alone once it reaches working.
-a
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
