Re: PE2950 and Linux virtualization

["Matt Shields-5" <user=70@nabble.blu.org>]

 On Thu, Jun 5, 2008 at 11:50 PM, Kristian Erik Hermansen 
<[hidden email]> wrote: 
> On Thu, Jun 5, 2008 at 5:55 PM, Scott R. Ehrlich <[hidden email]> wrote: 
>> What are people's experiences with CentOS 5.0 64-bit installed on dual 3 Ghz 
>> quad-core PE2950 systems with 32 GB RAM each, high-performance computing 
>> (applications that tax both the CPUs and RAM), not currently in a Beowolf 
>> cluster but could adapt to that, and doing so with VMWare or other 
>> vitualization software vs activity being done directly in the OS? 
> 
> I've done it.  The hardest past of clustering is properly 
> communicating state.  Check out erlang and/or Google's map-reduce to 
> understand how you can build a system which has very small tasks that 
> may be completed separately and then rejoined later to solve the 
> larger problem your aiming at... 
> 
>> How much of a performance hit, or gain (I'd presume hit), does 
>> virtualization cause an application, resulting in what percentage poorer or 
>> better (I'd presume poorer) performance vs dealing directly with the OS? 
> 
> About 6%, but even less with virtualization extensions... 
> 
>> It would be nice to have a VM perform some work, and if a person's code or 
>> application breaks, have it take down a VM while keeping a machine up, and 
>> not affecting other people's work. 
> 
> You want VMware's ESX platform (expensive) with VMotion... 
> http://www.vmware.com/products/vi/vc/vmotion.html
> 
>> It may also depend on if an application or code is written directly with/for 
>> the physical cpu/hardware vs more general use (VM). 
> 
> I don't know what you mean by this... 
> -- 
> Kristian Erik Hermansen 
> -- 
> "When you share your joys you double them; when you share your sorrows 
> you halve them." 
> 
> -- 
> This message has been scanned for viruses and 
> dangerous content by MailScanner, and is 
> believed to be clean.