Re: RAC or Large SMP...?
Date: Thu, 9 Oct 2008 18:12:36 -0500
Message-ID: <zdwHk.3145$c45.2962@nlpi065.nbdc.sbc.com>
<mccmx_at_hotmail.com> wrote in message
news:ca8eaaea-edf9-4c7f-a7b1-a04b6421fdfe_at_u40g2000pru.googlegroups.com...
>> Ok, just be aware though, if you don't use RAC at all, it mean downtime
>> if
>> the node crashes.- Hide quoted text -
>
> True. That is one big advantage of RAC, TAF.
>
> Once thing that I'm really curious about is this...
>
> If one of the 4 x3850 chassis died (hardware failure), could we
> restart the server with 3 chassis and continue to operate seemlessly
> (albeit with 75% of the processing. power)...?
>
According to page 197 of this document, you should be able to remove a node and restart.
http://www.redbooks.ibm.com/redpieces/pdfs/sg247630.pdf
> or would it require booting an alternate kernel, or modifying some
> boot/BIOS parameters..?
I would think that's the same as a single SMP box. If the primary boot device fails, we have to go to a mirror copy. Received on Thu Oct 09 2008 - 18:12:36 CDT