lkml.org 
[lkml]   [2000]   [Jan]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: Interesting analysis of linux kernel threading by IBM


Hi,

> As has been stated convincingly here (for me, at least) very high
> load means a runqueue of at the very most 10, ridiculous overload
> gives some 30. So your scheduler (using your own measurements!)
> hurts performance except for ridiculous overload (and contrieved
> benchmarks).

I couldn't agree more. I worked on two large AIX boxen this
year (8- and 12-way S7As with gigs of RAM) which were getting buried
under runqueues that ran past 2000 (no typo). The problem was not
in our software or in the kernel--it was simply very poor system
software design done by others in years past. There is nothing that
anyone can or should do to meet such loads; instead, one should look
at fixing the broken software design causing them.


Cheers,

Justin Fries
justinf@us.ibm.com
Raleigh, North Carolina



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:56    [W:0.023 / U:0.336 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site