SUMMARY: Cluster 3.1U4 event-logs

From: <Harald.Husemann_at_materna.de>
Date: Fri Dec 15 2006 - 07:42:03 EST
<original posting below>

Hello,

and thanks to Christopher Manly [cam2@cornell.edu] for his quick and reliable
answer! Here's his response:

===================/snip/=======================
These logs are primarily generated by GDS agents.  Somewhere along the
line, Sun changed the default logging on the GDS from silent to very
verbose.  As far as I can tell, these logs can only be read by a special
tool that Sun support only has access to.  So I decided to squelch
them.  The way to do that (aside from /dev/null'ing the logfiles) is to
set the Log_level extension property on GDS resources to 'NONE'.

Here's a script I wrote to do it across the cluster:

#!/bin/ksh

for group in `/usr/cluster/bin/scha_cluster_get -O ALL_RESOURCEGROUPS`
do
        for resource in `/usr/cluster/bin/scha_resourcegroup_get -O
RESOURCE_LIST -G $group`
        do
                isgds=`/usr/cluster/bin/scha_resource_get -O TYPE -R
$resource | /bin/cut -d : -f 1`

                if [ $isgds = 'SUNW.gds' ]
                then

Loglevel_info=`/usr/cluster/bin/scha_resource_get -O EXTENSION -R
$resource -G $group Log_level`
                        loglevel=`echo $Loglevel_info | /bin/awk '{print
$2}'`
                        #echo $resource $loglevel
                        if [ $loglevel != 'NONE' ]
                        then
                                /usr/cluster/bin/scrgadm -c -j $resource
-x Log_level=NONE
                                echo "$resource Fixed"
                        fi
                fi
        done
done


If you don't want to do that, play with this cron job in root's cron to
adjust the rate of rollover:

20 4 * * 0 /usr/cluster/lib/sc/newcleventlog /var/cluster/logs/eventlog

--
Christopher Manly
Unix Systems Administrator
CIT Systems & Operations
706 Rhodes Hall
Ithaca, NY 14853
607-255-0443
cam2@cornell.edu
===============/snap/=================================

I think I will reconfigure the log-level, as Christopher suggested, to get rid
of these logs.

Thanks to Christopher and to the list,

have a nice hackin',

Harald

Original posting:

> Hello gurus,
>
> can anyone out there shed some light on the event logs which are written
> in /var/cluster/logs? The problem is that these logs are growing quite
> fast, and are fillin' up my /var-partition.
> I'd like to make the logging less verbose, and have the cluster g'zip
> this logs automatically after rotating to save space. What is the
> recommended way to do this? (Of course I can use "find" to do the job,
> but I'd like to do it the "right" way, without introducing custom-made
> scripts)
> And, last not least, how can I read this logs? They seem to be
> binary-encoded, so...
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Fri Dec 15 07:42:35 2006

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:44:03 EST