SCIENTIFIC-LINUX-USERS Archives

January 2008

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Stephen John Smoogen <[log in to unmask]>
Reply To:
Stephen John Smoogen <[log in to unmask]>
Date:
Sun, 20 Jan 2008 22:07:25 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (27 lines)
On Jan 20, 2008 12:25 PM, markaoki <[log in to unmask]> wrote:
> The config limits chart for SL covers only V3 and 4, so I don't know for
> sure.
>
> Red Hat has a chart covering RHEL 3, 4, and 5, indicating a drop in
> supported
> maximum per-process memory, from 4gb to 3gb, for V5 over the prior versions:
> http://www.redhat.com/rhel/compare/
>
> I don't see hugemem kernel anywhere, not since SL 4 anyway.
>

I think the reason is that Red Hat is following Linus's lead on this:
If you want more than what the base architecture go to 64 bit. While
you can get 4gb on a 32 bit kernel, there is a non-negligible
performance hit and a lot of odd things seem to happen that are not
reproducible.. but thought to be due to the various hacks one has to
on the chipsets to get it to be more ram than it is supposed to.




-- 
Stephen J Smoogen. -- CSIRT/Linux System Administrator
How far that little candle throws his beams! So shines a good deed
in a naughty world. = Shakespeare. "The Merchant of Venice"

ATOM RSS1 RSS2