> Hi
>
> I have just setup a couple of windows server 2003 as virtual servers, one
> in MS Virtual Server R2, one in VmWare Server. They are running on a HP
> DL385 with one 2.4 Ghz dual-core Opteron and 5 GB of ram, system on
> mirrored 72Gb disks, virtual servers on 4 300Gb disks in Raid 1+0.
>
> Both are configured with 512MB ram and 16 Gb SCSI-disk, 1 cpu.
>
> I am trying to evaluate the performance of the virtual software. I have
> not been able to find any tests that uses R2, and have been irritated a
> bit by those "VmWare is faster than MS Virtual Server - period" statements
> you see by people who has not done any proper tests.
>
> So far I have tried running SiSoft Sandra and PCMark and have not been
> able to find any significant differences in performance. For example the
> SiSoft Sandra Multi-Media test, the CPU Arithmetic, the Memory Bandwith
> and others, they are within a few percents of each other. From what I can
> see in my quick tests, the variance of the tests are higher than the
> difference of the results between MS and VmWare.
>
> The only significant performance difference I can see is that VmWare
> Server can run 2 cpu in the virtual machine. That of course makes a
> difference.
>
> So, do anybody have any good tests that I could try on the virtual
> machines? I'm not going to spend days testing, but if anybody has a quick
> test I would like to run it to get som more numbers. I would like to do
> some more server-application oriented tests, but I can't seem to find any
> general tests that don't involve a big setup.
>
> Right now, the way I see it performance should not be the thing that
> decides between VmWare and MS (unless you got to have dual-cpu in the
> virtual machine (I don't, those I still prefer to run physical)).
>
> Regards,
>
> Kaj
>

> "Andrew Dugdell" <andrew @virtualserver.tv> wrote in message
> news:(email address - cut out)...
>>
>> I tried to keep the tests/taks "real world", and if you are doing the
>> same I would recommend turning off the debugging option under VMware
>> Server (it is enabled by default for VMs).
>
> Good hint - did not think about this. I will rerun some of the tests on
> VMware.
> Ok. I will run a few test with IOmeter to test harddisk performance, and
> might test a bit with Web Application Stress tool. I will keep newsgroup
> updated, and hopefully will you :-)

> Hi
>
> I have just setup a couple of windows server 2003 as virtual servers, one in
> MS Virtual Server R2, one in VmWare Server. They are running on a HP DL385
> with one 2.4 Ghz dual-core Opteron and 5 GB of ram, system on mirrored 72Gb
> disks, virtual servers on 4 300Gb disks in Raid 1+0.
>
> Both are configured with 512MB ram and 16 Gb SCSI-disk, 1 cpu.
>
> I am trying to evaluate the performance of the virtual software. I have not
> been able to find any tests that uses R2, and have been irritated a bit by
> those "VmWare is faster than MS Virtual Server - period" statements you see
> by people who has not done any proper tests.
>
> So far I have tried running SiSoft Sandra and PCMark and have not been able
> to find any significant differences in performance. For example the SiSoft
> Sandra Multi-Media test, the CPU Arithmetic, the Memory Bandwith and others,
> they are within a few percents of each other. From what I can see in my
> quick tests, the variance of the tests are higher than the difference of the
> results between MS and VmWare.
>
> The only significant performance difference I can see is that VmWare Server
> can run 2 cpu in the virtual machine. That of course makes a difference.
>
> So, do anybody have any good tests that I could try on the virtual machines?
> I'm not going to spend days testing, but if anybody has a quick test I would
> like to run it to get som more numbers. I would like to do some more
> server-application oriented tests, but I can't seem to find any general
> tests that don't involve a big setup.
>
> Right now, the way I see it performance should not be the thing that decides
> between VmWare and MS (unless you got to have dual-cpu in the virtual
> machine (I don't, those I still prefer to run physical)).
>
> Regards,
>
> Kaj
>

> Hi
>
> I have just setup a couple of windows server 2003 as virtual servers, one in
> MS Virtual Server R2, one in VmWare Server. They are running on a HP DL385
> with one 2.4 Ghz dual-core Opteron and 5 GB of ram, system on mirrored 72Gb
> disks, virtual servers on 4 300Gb disks in Raid 1+0.
>
> Both are configured with 512MB ram and 16 Gb SCSI-disk, 1 cpu.
>
> I am trying to evaluate the performance of the virtual software. I have not
> been able to find any tests that uses R2, and have been irritated a bit by
> those "VmWare is faster than MS Virtual Server - period" statements you see
> by people who has not done any proper tests.
>
> So far I have tried running SiSoft Sandra and PCMark and have not been able
> to find any significant differences in performance. For example the SiSoft
> Sandra Multi-Media test, the CPU Arithmetic, the Memory Bandwith and others,
> they are within a few percents of each other. From what I can see in my
> quick tests, the variance of the tests are higher than the difference of the
> results between MS and VmWare.
>
> The only significant performance difference I can see is that VmWare Server
> can run 2 cpu in the virtual machine. That of course makes a difference.
>
> So, do anybody have any good tests that I could try on the virtual machines?
> I'm not going to spend days testing, but if anybody has a quick test I would
> like to run it to get som more numbers. I would like to do some more
> server-application oriented tests, but I can't seem to find any general
> tests that don't involve a big setup.
>
> Right now, the way I see it performance should not be the thing that decides
> between VmWare and MS (unless you got to have dual-cpu in the virtual
> machine (I don't, those I still prefer to run physical)).
>
> Regards,
>
> Kaj
>