[Gllug] Crashes (owing to the heatwave?)

[David Damerell]

On Friday, 8 Aug 2003, Jon Masters wrote:
>On Fri, 8 Aug 2003, David Damerell wrote:
>>On a Sun E10K or later
>When I heard this support had been added to Solaris is stuck me as a bit
>of a real world bodge to give enough flexibility to seem partitioned.

What the marketing didn't mention is that the dynamic repartitioning
tools were like pulling teeth, and the "automatic" repartitioning
meant "you can schedule this via cron"; and that repartitioning
occasionally kills the partition.

Sun's partitioning was a step forward for them but it certainly is
nothing like what IBM have.

>My point was that if the CPU hotswap is done as part of the partitioning
>abstraction then most of the work does not happen inside Linux.

You're absolutely right (as far as I know _none_ of the work happens
inside Linux - the hosted OSes ideas of the number of CPUs and amount
of memory they have is not obliged to correspond to reality).

-- 
David Damerell <damerell at chiark.greenend.org.uk> Kill the tomato!

-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug