[K12OSN] Just about ready to make a purchase

[Robert Arkiletian]

On 5/9/07, Jim Kronebusch <jim at winonacotter.org> wrote:
> > Your post stuck in my head this evening Jim. Few more points I wanted
> > to share. Those 8 cores are fed by (share) one 800Mhz FSB memory
> > controller using UMA (Uniform Memory Access). This is where Opteron
> > systems shine. Each Opteron has it's own memory controller. So you
> > would have 4 memory controllers controlling 4 memory banks. However, I
> > am not sure if NUMA (Non Uniform Memory Access) in the Linux kernel
> > works with dual core cpu's. Maybe someone can chime in if they know.
> > But I know 4 single core cpu's would enable NUMA support.
>
> I have never heard of this.  I am trying to find information about this on Dell's site,
> no luck so far.  Everything I find suggests that RedHat Linux Enterprise AS can take
> full advantage of the multiple core multiprocessor systems, but they do not refer to
> NUMA.  Knowing that each processor will have full access to memory would be nice.

Each core does have access to all the memory but they have to wait
their turn since there is only one memory controller. Intel
compensates for this design by increasing the amount of cache on the
cpu.

>
> > BTW, 300GB scsi drives are awful expensive. If you are going to use
> > hardware RAID 10 have you considered enterprise SATA drives and a
> > dedicated PCI-X SATA controller with onboard write back cache (3ware
> > 9550SX or LSI MegaRAID 300-8X). My guess is it would probably be
> > better than the controller that comes with the Dell. Maybe someone who
> > has used those controllers can comment.
>
> yeah, very expensive.  But I have read so many things that say SCSI is still far
> superior to SATA for read/write speeds especially when needing to randomly access a lot
> of data.  Also I have seen information that the physical construction of a SCSI drive is
> still far superior to SATA.  I have given Hardware RAID with SATA a try on a few servers
> where I needed to save a customer money.  In my opinion those servers were slower than
> identical servers I configured with SCSI.  Since I can buy SCSI in my price range, I
> feel safer going with it.
>
> Do you really think that the SATA config you suggest would outperform the SCSI?  Or is
> this just a money saving suggestion?
>

Money saving suggestion.


> > Just checked the Dell PowerEdge 6800 specs and they have this under
> > the power supply section.
> >
> > "Redundant power is available in all 200-240V configurations and most
> > 100-119V / 120-127V configurations except those that contain or exceed
> > four processors, 32GB memory, five hard drives, and two PCI cards.
> > Redundant power availability will vary by configuration. NOTE: The
> > minimum configuration for most US based facilities is 120-127 Volts."
> >
> > According to this, your setup probably requires too much power to
> > support redundancy.
>
> I noticed that too.  They don't come right out and state that you can't use redundant,
> but they say their techs will look at the components and determine whether or not it
> will handle it.  I don't really want to switch to some 200-240v data center config, so I
> have basically conceded that it will not be redundant.  I always buy the 4hr response
> support from Dell and have had good luck with that.  If we go down for 4-6 hours once or
> twice a year even (which shouldn't happen), I don't think that would be that big of a
> deal.

This kind of concerns me. By having a dual PS, where you need both to
function to supply enough power, it effectively reduces your MTBF of
your PS. Not good. If Dell is selling dual redundant PS they should
ensure the system can work on only one PS. Remember even if you have a
battery backup it won't help if one of your PS dies.

-- 
Robert Arkiletian
Eric Hamber Secondary, Vancouver, Canada
Fl_TeacherTool http://www3.telus.net/public/robark/Fl_TeacherTool/
C++ GUI tutorial http://www3.telus.net/public/robark/