If you've followed any of our recent messages, you know we recently
crashed, and with the help of folks on the list(s), we have been able to
get back on-line and stable again.
We've enabled the SMP switch in the makefile, and now have both of our
processors spun up, and the kernel has remained perfectly agreeable since
booting it up two days ago.
My question involves load averages. When doing any intense disk I/O, with
either the adaptec 2940-based Seagate Hawk, or the Triton II-based WD
Caviar, the load average climbs dramatically; I've seen it go as high as
76.20!! and, believe it or not, the system didn't crash. Since then, I've
written my scripts to sleep at startegic points during backup, and niced
the embedded processes into oblivion, and the load avg only exceeds one at
a couple of points over the run of the backup script.
My question is this: what -is- a high load average? Over most of the
period of a year of running this system as a production fileserver, The
'average' load average is probably near .25; so just what -are- reasonable
loads on this system? (disk subsystems as described; dual P-166s, not
overclocked; 32MB of crappy ram (we're cursed with the SIG-11 and so we
compile off of this machine) on an older Tyan II Tomcat.
Also, we're running a 2.0.29 kernel, with no patches. Any patches anyone
would like to suggest would be greatly appreciated.
-James
...
He who knows others is wise.
He who knows himself is enlightened.
-- Lao Tsu