> Can anyone spot where the "weak link" is
in my system. I put in this
> post (what I believe) is the most relevant information...
judging from the huge duration of these timeouts
it really doesn't look like you are running a rt
kernel.
uname -a
says?
bash# uname -a
Linux localhostlocaldomain 2.6.16-1.2080.16.rrt.rhfc5.ccrma #1 PREEMPT
Tue Jul 25 15:06:37 EDT 2006 i686 i686 i386 GNU/Linux
Maybe I'm confused about what it really means to run an rt kernel, but
I'm running FC5/CCRMA and this is what the install page reads...
========================================================
Installing the low latency kernel
Planet CCRMA provides versions of the Linux kernel with one crucial
difference with the standard kernel, the inclusion of the realtime
preemption patch by Ingo Molnar.
Planet CCRMA on Fedora Core 5 currently includes two kernels with
slightly different configurations of the realtime preemption patch. The
more conservative is configured as PREEMPT_DESKTOP, the more adventurous
is configured as PREEMPT_RT. The latter will give you (if it works on
your hardware) much better low latency performance. As usual what you
get in terms of latency depends on the exact mix of hardware you are
running on.
========================================================
I'm running the kernel configured as PREEMPT_RT.
also I found that the rtirq script was "not doing
the right thing"
on later kernels
Also give ingo's kernel a shot
I'll have to check with the CCRMA list and see if anyone has had
problems with the rtirq script on their FC5/CCRMA systems with this kernel.
Is cpuspeed running? If so, disable it.
Bear with me... this is all new to me...
Do I understand that 'chkconfig' would be the command to disable
cpuspeed? If so, would you suggest that I disable it for all runlevels?
This is the output of '/sbin/chkconfig --list cpuspeed'...
cpuspeed 0:off 1:on 2:on 3:on 4:on 5:on 6:off
So... '/sbin/chkconfig --level 12345 cpuspeed off' would disable it for
all runlevels?
Rocco