Hello,
running NOCOL I've been getting a critical error message regarding
the number og PageIns of one of our Linux boxes. The system in question
is running a Linux 2.0.35 kernel.
It would appear that the number of PageIns and PageOuts is cleaned from
/proc/stat by hostmon-osclient.linux. However, whereas NOCOL is assuming
PageIns / Second (since it takes these values from vmstat on a Solaris
system and the Solaris vmstat man page states that these values are given
as pi/s), according to the man page on /proc for Linux in stat we apparently
find the accumulated PageIns since system boot. Thus, after a certain uptime,
a Linux system will always get a critical error on PageIns.
Any ideas how to work around this and still get warned if PageIns are
excessive?
Thanx in advance,
Martin
--
Martin Boening, MB3792 | EMail: mboen@sbs.de
SBS SCN22, Otto-Hahn-Ring 6, D-81739 Muenchen (Perlach), Germany
Phone: +49 896 364 2904 FAX: +49 525 183 1070
As far as we know, our computer has never had an undetected error.
-- Weisert
|