diff options
Diffstat (limited to 'contrib/aix')
| -rw-r--r-- | contrib/aix/check_crit_dsk | 66 | ||||
| -rw-r--r-- | contrib/aix/check_dsk | 62 | ||||
| -rw-r--r-- | contrib/aix/check_failed | 48 | ||||
| -rw-r--r-- | contrib/aix/check_io | 69 | ||||
| -rw-r--r-- | contrib/aix/check_kerberos | 49 | ||||
| -rw-r--r-- | contrib/aix/check_ping | 117 | ||||
| -rw-r--r-- | contrib/aix/check_queue | 67 | ||||
| -rw-r--r-- | contrib/aix/pg_stat | 45 |
8 files changed, 523 insertions, 0 deletions
diff --git a/contrib/aix/check_crit_dsk b/contrib/aix/check_crit_dsk new file mode 100644 index 00000000..566e07c1 --- /dev/null +++ b/contrib/aix/check_crit_dsk @@ -0,0 +1,66 @@ +#!/bin/sh + +#========================================================================= +# Critical Disk Checker utility +# +# This is the same as the disk checker utility but we use it as +# a seperate service in Nagios to report on partitions that +# have reached 100% capacity. +# +# We have excluded /dev/cd0 because the cdrom drive will always +# report 100% capacity if a CD is in the drive. +# +# Authors: TheRocker +# SpEnTBoY +# +# Email: therocker@pawprints.2y.net +# lonny@abyss.za.org +# +#======================================================================= + +NUMBER=`rsh $1 -l root df -kP | grep -vE ":|/dev/cd0" | grep -E "100%" | tr -s ' '| cut -d' ' -f5 | cut -c1-3 | line` +TMPFILE=/tmp/tmpcrit.hndl +TMPTOO=/tmp/twocrit.hndl + +if [ "$NUMBER" -eq 100 ] +then + + `rsh $1 -l root df -kP |grep -vE ":|/dev/cd0" | grep -E "100%" | tr -s ' '| cut -d' ' -f6,5 >> $TMPFILE` + + LINES=`wc -l /tmp/tmpcrit.hndl | cut -c8` + LINESCTL=`wc -l /tmp/tmpcrit.hndl | cut -c8 ` + echo "Filesystems over 99% --> \c" + +#=============================================================== +# Just a little bit to check for multiple occurances of the +# condition. +#=============================================================== + + while [ $LINESCTL != 0 ] + do + + cat $TMPFILE | tail -$LINESCTL > $TMPTOO + cat $TMPTOO > $TMPFILE + LINESCTL=$(( $LINESCTL -1 )) + LINES=$(( $LINES -1 )) + DATA=`head -1 /tmp/tmpcrit.hndl` + echo "( $DATA ) \c" + + + done + echo "\n" + +#=============================================================== +# File clean up. Always pick up after yourself. Disk space +# doesn't grow on trees you know. +#=============================================================== + + rm -f $TMPFILE + rm -f $TMPTOO + exit 2 + +else + + echo "No Filesystems over 99%... OK" + exit 0 +fi diff --git a/contrib/aix/check_dsk b/contrib/aix/check_dsk new file mode 100644 index 00000000..c8ddb3f8 --- /dev/null +++ b/contrib/aix/check_dsk @@ -0,0 +1,62 @@ +#! /bin/sh + +#====================================================================== +# Disk Checker utility +# +# Simple little script that checks the status of all partitions +# on a node's hard disks. It will produce a warning alert and list +# the offending filesystems in nagios. +# +# Authors: SpEnTBoY +# TheRocker +# +# Email: lonny@abyss.za.org +# therocker@pawprints.2y.net +#===================================================================== + +NUMBER=`rsh $1 -l root df -kP | grep -v ":" | grep -E "9[0-9]%" | tr -s ' '| cut -d' ' -f5 | cut -c1-2 | line` +TMPFILE=/tmp/tmp.hndl +TMPTOO=/tmp/two.hndl + +if [ "$NUMBER" -gt 90 ] +then + + `rsh $1 -l root df -kP |grep -v ":" | grep -E "9[0-9]%" | tr -s ' '| cut -d' ' -f6,5 >> $TMPFILE` + + LINES=`wc -l /tmp/tmp.hndl | cut -c8` + LINESCTL=`wc -l /tmp/tmp.hndl | cut -c8 ` + echo "Filesystems over 90% --> \c" + +#====================================================================== +# You'll see this one in a few our shell scripts. Just chcecking for +# multiple occurances of the warnign condition. We gotta list 'em all +#====================================================================== + + while [ $LINESCTL != 0 ] + do + + cat $TMPFILE | tail -$LINESCTL > $TMPTOO + cat $TMPTOO > $TMPFILE + LINESCTL=$(( $LINESCTL -1 )) + LINES=$(( $LINES -1 )) + DATA=`head -1 /tmp/tmp.hndl` + echo "( $DATA ) \c" + + + done + echo "\n" + +#=============================================================== +# Clean up all those nasty tmp files that suck up valuable +# disk realestate. +#=============================================================== + + rm -f $TMPFILE + rm -f $TMPTOO + exit 1 + +else + + echo "No Filesystems over 90%... OK" + exit 0 +fi diff --git a/contrib/aix/check_failed b/contrib/aix/check_failed new file mode 100644 index 00000000..50cdf7e1 --- /dev/null +++ b/contrib/aix/check_failed @@ -0,0 +1,48 @@ +#!/usr/bin/perl +#====================== +# Created May 25, 2000 +#====================== + +# This scripts is for checking for failed root login attempts on +# any machine running AIX which has a failedlogin file in /etc/security +# The purpose is to thwart (good word) any unauthorised people from +# even trying to log in as root. This plugin has been developed for Nagios +# running on AIX. +# Lonny Selinger SpEnTBoY lonny@abyss.za.org +# May + + +my $server = $ARGV[0]; + +if (!$ARGV[0]) { + print "You must specify a server to check\n"; + print "usage: ./check_failed <Server Name>\n"; + exit (-1); + } else { + open (DATE, "/bin/date '+%b %d' |"); + while (<DATE>) { + $dline = $_; + @dresults = $dline; + chop $dresults[0]; + } + open (SULOG, "rsh $server -l root who /etc/security/failedlogin | grep root |"); + while (<SULOG>) { + $line = $_; + @results = split (/\s+/,$line); + if ($line =~ /^root/) { + if (join(' ', @results[2,3]) eq $dresults[0]) { + print "FAILED root login on $dresults[0], node: $ARGV[0] from $results[5]\n"; + exit(2); + } + } + } +} +if (join(' ', @results[2,3]) ne $dresults[0]) { + print "No Failed Root Logins on This Node\n"; + exit(0); +} +exit(0); +close(SULOG); +close(DATE); + + diff --git a/contrib/aix/check_io b/contrib/aix/check_io new file mode 100644 index 00000000..58b25f6d --- /dev/null +++ b/contrib/aix/check_io @@ -0,0 +1,69 @@ +#! /bin/sh + +#================================================================= +# +# I/O Checker (KBPS) +# This Script uses iostat to monitor disk io +# Useful for notifications of disk thrashing. +# +# Authors: TheRocker +# SpEnTBoY +# +# Email: therocker@pawprints.2y.net +# lonny@abyss.za.org +# +#================================================================ + +NUMBER1=`rsh $1 -l root iostat -d | grep -e "hdisk" | tr -s ' ' | cut -d' ' -f2 | sort -2 -r | cut -c1 | line` +NUMBER2=`rsh $1 -l root iostat -d | grep -e "hdisk" | tr -s ' ' | cut -d' ' -f2 | sort -2 -r | cut -c2 | line` +TMPFILE=/tmp/iotest.hndl +TMPTOO=/tmp/iotwo.hndl + +#=========================================================== +# +# We do an evaluation on $NUMBER1 and $NUMBER2 to see if +# disk io is exceeding 40%. +# +#=========================================================== + +if [ "$NUMBER1" -gt 4 ] && [ "$NUMBER2" -gt 0 ] +then + + `rsh $1 -l root iostat -d | grep -v cd0 | tr -s ' '| cut -d' ' -f1,2 | grep -e "4[0-9]." >> $TMPFILE` + +#==================================================================== +# +# Of course, there may be more than one hard disk on the node +# so we use this bit of code to report on more than one instance +# of excessive disk IO. +# +#==================================================================== + + LINES=`wc -l /tmp/iotest.hndl | cut -c8` + LINESCTL=`wc -l /tmp/iotest.hndl | cut -c8 ` + echo "WARNING!!! Disk I/O Exceeding 40% on --> \c" + + while [ $LINESCTL != 0 ] + do + + cat $TMPFILE | tail -$LINESCTL > $TMPTOO + cat $TMPTOO > $TMPFILE + LINESCTL=$(( $LINESCTL -1 )) + LINES=$(( $LINES -1 )) + DATA=`head -1 /tmp/iotest.hndl` + echo "( $DATA ) " + + + done + echo "\n" + + rm -f $TMPFILE + rm -f $TMPTOO + exit 1 + +else + + print "No Disk I/O Exceeding 40%...OK" + exit 0 + +fi diff --git a/contrib/aix/check_kerberos b/contrib/aix/check_kerberos new file mode 100644 index 00000000..443ab109 --- /dev/null +++ b/contrib/aix/check_kerberos @@ -0,0 +1,49 @@ +#! /bin/sh + +#========================================================================= +# Kerberos Ticket Checker +# +# This script is handy if you allow kerberos tickets to expire +# on your nodes. The script will simply warn you when a node has +# kerberos tickets expiring on the current date. This will allow to +# re-initialize the tickets if you wish to do so. +# +# Nothing fancy here, all Nagios will show is the number of tickets +# that are going to (or already have) expired. +# +# An item of note: +# +# We made no provisions for the weekend. If tickets expire on the +# weekend and nobody is around, you won't see a warning on the +# Nagios console because we look for expired on the current day +# only. It's a good idea to have this warning emailed to the +# appropriate admin and if there is something critical that relies +# on Kerberos, you might want to send a page. +# +# Authors: TheRocker +# SpEnTBoY +# +# Email: therocker@pawprints.2y.net +# lonny@abyss.za.org +#========================================================================= + +TMPFILE=/tmp/kerbtmp.hndl +DATE=`date +%b' '%d` + +rsh $1 -l root /usr/lpp/ssp/kerberos/bin/klist | tr -s ' ' | cut -d' ' -f4,5,6 | grep -e "$DATE" > $TMPFILE + + +if [ -s $TMPFILE ] +then + + LINES=`wc -l /tmp/kerbtmp.hndl | cut -c7-8` + echo "Kerberos Tickets set to expire --> \c" + echo "$LINES \c" + echo "\n" + + rm -f $TMPFILE + exit 1 + +fi + echo "Kerberos Tickets are valid" + exit 0 diff --git a/contrib/aix/check_ping b/contrib/aix/check_ping new file mode 100644 index 00000000..aaa8c84e --- /dev/null +++ b/contrib/aix/check_ping @@ -0,0 +1,117 @@ +#!/usr/bin/perl -w + +#================================================================ +# +# This perl script will accept an argument and simply pass it +# to ping. It works by sending 2 ping to the specified host +# and evaluating on the average delta time of those 2 pings. +# +# Author: SpEnTBoY +# Email: lonny@abyss.za.org +# April 5,2000 +# +#================================================================ + +#============================ +# State predefined stuff and +# requirements +#============================ + +require 5.004; +use POSIX; +use strict; + +sub usage; + +my $ipaddr = $ARGV[0]; + +my $TIMEOUT = 15; + +my %ERRORS = ('UNKNOWN' , '-1', + 'OK' , '0', + 'WARNING', '1', + 'CRITICAL', '2'); + +my $remote = shift || &usage(%ERRORS); +my $warning = shift || 750; +my $critical = shift || 1000; + +my $state = "OK"; +my $answer = undef; +my $offset = undef; +my $line = undef; + +#============================================================ +# If theres no response we can exit the bloody thing cleanly +# last thing I want to do is hang an AIX system ;-) +#============================================================ + +$SIG{'ALRM'} = sub { + print ("ERROR: No response from PING! (alarm)\n"); + exit $ERRORS{"UNKNOWN"}; +}; +alarm($TIMEOUT); + +#================================================ +# Pass stddn from $ARGV to the command and parse +# the info we need (namely the value for "max" +#================================================ + + + +open(PING,"/usr/sbin/ping -c 2 '$ipaddr' >&1|"); +while (<PING>) { + $line = $_; + if (/round-trip min\/avg\/max = (.+)\/(.+)\/(.+) ms/) { + $offset = $3; + last; + } +} + +#================================================== +# Do some error checking on the output of the file +# and implement values for <crit> and <warn> +# deffinitions if they were specified by the user +# or sub in the predefined ones +#================================================== + +if (defined $offset) { + if (abs($offset) > $warning) { + if (abs($offset) > $critical) { + $state = "CRITICAL"; + $answer = ": Ping Time $offset MS greater than +/- $critical MS\n"; + } else { + $state = "WARNING"; + $answer = ": Ping Time $offset MS greater than +/- $warning MS\n"; + } + } else { + $state = "OK"; + $answer = ": Ping Time $offset MS\n"; + } +} else { + $state = "UNKNOWN"; + $answer = ": $line\n"; +} +print ("$state$answer"); +exit $ERRORS{$state}; + +sub usage { + print "\n"; + print "#=========================================\n"; + print "Check_Ping 0.02 script by Lonny Selinger\n"; + print "Made with AIX in mind ;-)\n"; + print "#=========================================\n"; + print "\n"; + print "#================================================\n"; + print " I'm going to need a few more arguments from you\n"; + print "#================================================\n"; + print "\n"; + print "#================================================\n"; + print "Usage: check_ping <host> [<warn> [<crit>]\n"; + print "#================================================\n"; + print "\n"; + print "<warn> = Ping in MS at which a warning message will be generated.\n Defaults to 750.\n"; + print "<crit> = Ping in MS at which a critical message will be generated.\n Defaults to 1000.\n\n"; + exit $ERRORS{"UNKNOWN"}; +} + diff --git a/contrib/aix/check_queue b/contrib/aix/check_queue new file mode 100644 index 00000000..9f709c54 --- /dev/null +++ b/contrib/aix/check_queue @@ -0,0 +1,67 @@ +#! /bin/sh + +#=============================================================== +# Print Queue Checker +# +# The print queue checker simply looks for an occurance of a +# DOWN queue. A note of warning, if you use remote queues in +# AIX to redirect print jobs from the AIX queue to an NT print +# server that print through DLC rather than IP, it will be very +# s - l - o - w. But it will work. +# +# Author: TheRocker +# Email: therocker@pawprints.2y.net +#=============================================================== + +TMPFILE=/tmp/qtmp.hndl +TMPTOO=/tmp/qtwo.hndl + +#======================================================================= +# +# This script will also work on AIX 4.2.1 BUT you have to change +# the following line. AIX 4.2.1 does not support the -W option +# with lpstat. For AIX 4.2.1 just remove the -W option and it should +# work just fine. +# +#======================================================================= + +`rsh $1 -l root lpstat -W | grep -e "DOWN" | tr -s ' ' | cut -d' ' -f1,3 > /tmp/qtmp.hndl 2> /tmp/q_err` + +if [ -s $TMPFILE ] +then + +#======================================================= +# +# If you've seen the other AIX scripts I wrote you may +# notice that I use this bit of code a lot. Well it +# works and appears to be all purpose. +# +#======================================================= + + LINES=`wc -l /tmp/qtmp.hndl | cut -c8` + LINESCTL=`wc -l /tmp/qtmp.hndl | cut -c8` + + echo "Print Queue DOWN --> \c" + + while [ $LINESCTL != 0 ] + do + + cat $TMPFILE | tail -$LINESCTL > $TMPTOO + cat $TMPTOO > $TMPFILE + LINESCTL=$(( $LINESCTL -1 )) + LINES=$(( $LINES -1 )) + DATA=`head -1 /tmp/qtmp.hndl` + echo "( $DATA ) \c" + + + done + + echo "\n" + + rm -f $TMPFILE + rm -f $TMPTOO + exit 2 + +fi + echo "Print Queues Running... OK" + exit 0 diff --git a/contrib/aix/pg_stat b/contrib/aix/pg_stat new file mode 100644 index 00000000..e0603ec4 --- /dev/null +++ b/contrib/aix/pg_stat @@ -0,0 +1,45 @@ +#!/bin/ksh + +#============================================================================== +# Script was originally created to collect stats and dump then to a log file +# every five minutes. But we like this better (the log file thing is still +# good if you want to track availability). +# +# Authors: SpEnTBoY +# TheRocker +# +# Email: lonny@abyss.za.org +# therocker@pawprints.2y.net +#============================================================================== + +#========================================================================================= +# +# The best way to do this is to use Kerberos but we use rsh here because our monitoring +# workstation doesn't have Kerberos installed. In order for this to work, the remote +# host ($1) must have a .rhosts file that contains a line like: +# +# monitorhost nagiosuser +# +#========================================================================================= + +PAGING2=`rsh $1 -l root lsps -a -s | grep -v Paging | tr -s ' '| cut -d' ' -f3 | cut -d'%' -f1` + + +if [ "$PAGING2" -gt "35" ] && [ "$PAGING2" -lt "50" ] +then + echo "Paging Space is over 35% ("$PAGING2")%" +exit 1 +fi + +if [ "$PAGING2" -gt "49" ] +then + echo "WARNING! Paging Space is over 50% ("$PAGING2")%" +exit 2 +fi + +if [ "$PAGING2" -lt "34" ] +then + echo "Paging Space is less than 34% ("$PAGING2")%" +exit 0 +fi + |