Skip to content

Instantly share code, notes, and snippets.

@lunetics
Created April 27, 2015 21:07
Show Gist options
  • Select an option

  • Save lunetics/a064ecfce8b816cb261e to your computer and use it in GitHub Desktop.

Select an option

Save lunetics/a064ecfce8b816cb261e to your computer and use it in GitHub Desktop.

Revisions

  1. lunetics created this gist Apr 27, 2015.
    437 changes: 437 additions & 0 deletions check_mk_agent.freebsd
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,437 @@
    #!/usr/bin/env bash
    # +------------------------------------------------------------------+
    # | ____ _ _ __ __ _ __ |
    # | / ___| |__ ___ ___| | __ | \/ | |/ / |
    # | | | | '_ \ / _ \/ __| |/ / | |\/| | ' / |
    # | | |___| | | | __/ (__| < | | | | . \ |
    # | \____|_| |_|\___|\___|_|\_\___|_| |_|_|\_\ |
    # | |
    # | Copyright Mathias Kettner 2014 [email protected] |
    # +------------------------------------------------------------------+
    #
    # This file is part of Check_MK.
    # The official homepage is at http://mathias-kettner.de/check_mk.
    #
    # check_mk is free software; you can redistribute it and/or modify it
    # under the terms of the GNU General Public License as published by
    # the Free Software Foundation in version 2. check_mk is distributed
    # in the hope that it will be useful, but WITHOUT ANY WARRANTY; with-
    # out even the implied warranty of MERCHANTABILITY or FITNESS FOR A
    # PARTICULAR PURPOSE. See the GNU General Public License for more de-
    # ails. You should have received a copy of the GNU General Public
    # License along with GNU Make; see the file COPYING. If not, write
    # to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
    # Boston, MA 02110-1301 USA.

    # Author: Lars Michelsen <[email protected]>
    # Florian Heigl <[email protected]>
    # (Added sections: df mount mem netctr ipmitool)

    # NOTE: This agent has beed adapted from the Check_MK linux agent.
    # The most sections are commented out at the moment because
    # they have not been ported yet. We will try to adapt most
    # sections to print out the same output as the linux agent so
    # that the current checks can be used.

    # This might be a good source as description of sysctl output:
    # http://people.freebsd.org/~hmp/utilities/satbl/_sysctl.html

    # Remove locale settings to eliminate localized outputs where possible
    export LC_ALL=C
    unset LANG

    export MK_LIBDIR="/usr/local/lib/check_mk_agent"
    export MK_CONFDIR="/etc/check_mk"
    export MK_TMPDIR="/var/run/check_mk"


    # Make sure, locally installed binaries are found
    PATH=$PATH:/usr/local/bin:/usr/local/sbin

    # All executables in PLUGINSDIR will simply be executed and their
    # ouput appended to the output of the agent. Plugins define their own
    # sections and must output headers with '<<<' and '>>>'
    PLUGINSDIR=$MK_LIBDIR/plugins

    # All executables in LOCALDIR will by executabled and their
    # output inserted into the section <<<local>>>. Please refer
    # to online documentation for details.
    LOCALDIR=$MK_LIBDIR/local


    # close standard input (for security reasons) and stderr
    if [ "$1" = -d ]
    then
    set -xv
    else
    exec </dev/null 2>/dev/null
    fi

    # Runs a command asynchronous by use of a cache file
    function run_cached() {
    if [ "$1" = -s ] ; then local section="echo '<<<$2>>>' ; " ; shift ; fi
    local NAME=$1
    local MAXAGE=$2
    shift 2
    local CMDLINE="$section$@"

    if [ ! -d $MK_TMPDIR/cache ]; then mkdir -p $MK_TMPDIR/cache ; fi
    CACHEFILE="$MK_TMPDIR/cache/$NAME.cache"

    # Check if the creation of the cache takes suspiciously long and return
    # nothing if the age (access time) of $CACHEFILE.new is twice the MAXAGE
    local NOW=$(date +%s)
    if [ -e "$CACHEFILE.new" ] ; then
    local CF_ATIME=$(stat -f "%a" "$CACHEFILE.new")
    if [ $((NOW - CF_ATIME)) -ge $((MAXAGE * 2)) ] ; then
    return
    fi
    fi

    # Check if cache file exists and is recent enough
    if [ -s "$CACHEFILE" ] ; then
    local MTIME=$(stat -f "%m" "$CACHEFILE")
    if [ $((NOW - MTIME)) -le $MAXAGE ] ; then local USE_CACHEFILE=1 ; fi
    # Output the file in any case, even if it is
    # outdated. The new file will not yet be available
    cat "$CACHEFILE"
    fi

    # Cache file outdated and new job not yet running? Start it
    if [ -z "$USE_CACHEFILE" -a ! -e "$CACHEFILE.new" ] ; then
    echo "$CMDLINE" | daemon bash -o noclobber > $CACHEFILE.new && mv $CACHEFILE.new $CACHEFILE || rm -f $CACHEFILE $CACHEFILE.new &
    fi
    }

    echo '<<<check_mk>>>'
    echo Version: 1.2.7i1
    echo AgentOS: freebsd
    echo PluginsDirectory: $PLUGINSDIR
    echo LocalDirectory: $LOCALDIR
    echo AgentDirectory: $MK_CONFDIR



    osver="$(uname -r)"
    is_jailed="$(sysctl -n security.jail.jailed)"


    # Partitionen (-P verhindert Zeilenumbruch bei langen Mountpunkten)
    # Achtung: NFS-Mounts werden grundsaetzlich ausgeblendet, um
    # Haenger zu vermeiden. Diese sollten ohnehin besser auf dem
    # Server, als auf dem Client ueberwacht werden.

    echo '<<<df>>>'
    # no special zfs handling so far, the ZFS.pools plugin has been tested to
    # work on FreeBSD
    if df -T > /dev/null ; then
    df -kTP -t ufs | egrep -v '(Filesystem|devfs|procfs|fdescfs|basejail)'
    else
    df -kP -t ufs | egrep -v '(Filesystem|devfs|procfs|fdescfs|basejail)' | awk '{ print $1,"ufs",$2,$3,$4,$5,$6 }'
    fi

    # Filesystem usage for ZFS
    if type zfs > /dev/null 2>&1 ; then
    echo '<<<zfsget>>>'
    zfs get -Hp name,quota,used,avail,mountpoint,type -t filesystem,volume || \
    zfs get -Hp name,quota,used,avail,mountpoint,type
    echo '[df]'
    df -kP -t zfs | sed 1d
    fi

    # Check NFS mounts by accessing them with stat -f (System
    # call statfs()). If this lasts more then 2 seconds we
    # consider it as hanging. We need waitmax.
    #if type waitmax >/dev/null
    #then
    # STAT_VERSION=$(stat --version | head -1 | cut -d" " -f4)
    # STAT_BROKE="5.3.0"
    #
    # echo '<<<nfsmounts>>>'
    # sed -n '/ nfs /s/[^ ]* \([^ ]*\) .*/\1/p' < /proc/mounts |
    # while read MP
    # do
    # if [ $STAT_VERSION != $STAT_BROKE ]; then
    # waitmax -s 9 2 stat -f -c "$MP ok %b %f %a %s" "$MP" || \
    # echo "$MP hanging 0 0 0 0"
    # else
    # waitmax -s 9 2 stat -f -c "$MP ok %b %f %a %s" "$MP" && \
    # printf '\n'|| echo "$MP hanging 0 0 0 0"
    # fi
    # done
    #fi

    # Check mount options.
    # FreeBSD doesn't do remount-ro on errors, but the users might consider
    # security related mount options more important.
    echo '<<<mounts>>>'
    mount -p -t ufs

    # processes including username, without kernel processes
    echo '<<<ps>>>'
    COLUMNS=10000
    if [ "$is_jailed" = "0" ]; then
    ps ax -o state,user,vsz,rss,pcpu,command | sed -e 1d -e '/\([^ ]*J\) */d' -e 's/ *\([^ ]*\) *\([^ ]*\) *\([^ ]*\) *\([^ ]*\) *\([^ ]*\) */(\2,\3,\4,\5) /'
    else
    ps ax -o user,vsz,rss,pcpu,command | sed -e 1d -e 's/ *\([^ ]*\) *\([^ ]*\) *\([^ ]*\) *\([^ ]*\) */(\1,\2,\3,\4) /'
    fi


    # Produce compatible load/cpu output to linux agent. Not so easy here.
    echo '<<<cpu>>>'
    echo `sysctl -n vm.loadavg | tr -d '{}'` `top -b -n 1 | grep -E '^[0-9]+ processes' | awk '{print $3"/"$1}'` `sysctl -n kern.lastpid` `sysctl -n hw.ncpu`

    # Calculate the uptime in seconds since epoch compatible to /proc/uptime in linux
    echo '<<<uptime>>>'
    up_seconds=$(( `date +%s` - `sysctl -n kern.boottime | cut -f1 -d\, | awk '{print $4}'`))
    idle_seconds=$(ps axw | grep idle | grep -v grep | awk '{print $4}' | cut -f1 -d\: )

    # second value can be grabbed from "idle" process cpu time / num_cores
    echo "$idle_seconds $up_seconds"


    # Platten- und RAID-Status von LSI-Controlleren, falls vorhanden
    #if which cfggen > /dev/null ; then
    # echo '<<<lsi>>>'
    # cfggen 0 DISPLAY | egrep '(Target ID|State|Volume ID|Status of volume)[[:space:]]*:' | sed -e 's/ *//g' -e 's/:/ /'
    #fi


    # Multipathing is supported in FreeBSD by now
    # http://www.mywushublog.com/2010/06/freebsd-and-multipath/
    if kldstat -v | grep g_multipath > /dev/null ; then
    echo '<<<freebsd_multipath>>>'
    gmultipath status | grep -v ^Name
    fi


    # Soft-RAID
    echo '<<<freebsd_geom_mirrors>>>'
    gmirror status | grep -v ^Name

    # Performancecounter Kernel
    echo "<<<kernel>>>"
    date +%s
    forks=`sysctl -n vm.stats.vm.v_forks`
    vforks=`sysctl -n vm.stats.vm.v_vforks`
    rforks=`sysctl -n vm.stats.vm.v_rforks`
    kthreads=`sysctl -n vm.stats.vm.v_kthreads`
    echo "cpu" `sysctl -n kern.cp_time | awk ' { print $1" "$2" "$3" "$5" "$4 } '`
    echo "ctxt" `sysctl -n vm.stats.sys.v_swtch`
    echo "processes" `expr $forks + $vforks + $rforks + $kthreads`

    # Network device statistics (Packets, Collisions, etc)
    # only the "Link/Num" interface has all counters.
    echo '<<<netctr>>>'
    date +%s
    if [ "$(echo $osver | cut -f1 -d\. )" -gt "8" ]; then
    netstat -inb | egrep -v '(^Name|lo|plip)' | grep Link | awk '{print $1" "$8" "$5" "$6" "$7" 0 0 0 0 "$11" "$9" "$10" 0 0 0 0 0"}'
    else
    # pad output for freebsd 7 and before
    netstat -inb | egrep -v '(^Name|lo|plip)' | grep Link | awk '{print $1" "$7" "$5" "$6" 0 0 0 0 0 "$10" "$8" "$9" 0 0 "$11" 0 0"}'
    fi


    # IPMI-Data (Fans, CPU, temperature, etc)
    # needs the sysutils/ipmitool and kldload ipmi.ko
    if which ipmitool >/dev/null ; then
    echo '<<<ipmi>>>'
    ipmitool sensor list \
    | grep -v 'command failed' \
    | sed -e 's/ *| */|/g' -e "s/ /_/g" -e 's/_*$//' -e 's/|/ /g' \
    | egrep -v '^[^ ]+ na ' \
    | grep -v ' discrete '
    fi


    # State of LSI MegaRAID controller via MegaCli.
    # To install: pkg install megacli
    if which MegaCli >/dev/null ; then
    echo '<<<megaraid_pdisks>>>'
    MegaCli -PDList -aALL -NoLog < /dev/null | egrep 'Enclosure|Raw Size|Slot Number|Device Id|Firmware state|Inquiry'
    echo '<<<megaraid_ldisks>>>'
    MegaCli -LDInfo -Lall -aALL -NoLog < /dev/null | egrep 'Size|State|Number|Adapter|Virtual'
    echo '<<<megaraid_bbu>>>'
    MegaCli -AdpBbuCmd -GetBbuStatus -aALL -NoLog < /dev/null | grep -v Exit
    fi


    # OpenVPN Clients.
    # Correct log location unknown, sed call might also be broken
    if [ -e /var/log/openvpn/openvpn-status.log ] ; then
    echo '<<<openvpn_clients:sep(44)>>>'
    sed -n -e '/CLIENT LIST/,/ROUTING TABLE/p' < /var/log/openvpn/openvpn-status.log | sed -e 1,3d -e '$d'
    fi


    if which ntpq > /dev/null 2>&1 ; then
    echo '<<<ntp>>>'
    # remote heading, make first column space separated
    ntpq -np | sed -e 1,2d -e 's/^\(.\)/\1 /' -e 's/^ /%/'
    fi


    # Checks for cups monitoring
    #if which lpstat > /dev/null 2>&1; then
    # echo '<<<cups_queues>>>'
    # lpstat -p
    # echo '---'
    # for i in $(lpstat -p | grep -E "^(printer|Drucker)" | awk '{print $2}' | grep -v "@"); do
    # lpstat -o "$i"
    # done
    #fi

    # Heartbeat monitoring
    #if which cl_status > /dev/null 2>&1; then
    # # Different handling for heartbeat clusters with and without CRM
    # # for the resource state
    # if [ -S /var/run/heartbeat/crm/cib_ro ]; then
    # echo '<<<heartbeat_crm>>>'
    # crm_mon -1 -r | grep -v ^$ | sed 's/^\s/_/g'
    # else
    # echo '<<<heartbeat_rscstatus>>>'
    # cl_status rscstatus
    # fi
    #
    # echo '<<<heartbeat_nodes>>>'
    # for NODE in $(cl_status listnodes); do
    # if [ $NODE != $HOSTNAME ]; then
    # STATUS=$(cl_status nodestatus $NODE)
    # echo -n "$NODE $STATUS"
    # for LINK in $(cl_status listhblinks $NODE 2>/dev/null); do
    # echo -n " $LINK $(cl_status hblinkstatus $NODE $LINK)"
    # done
    # echo
    # fi
    # done
    #fi

    # Number of TCP connections in the various states
    echo '<<<tcp_conn_stats>>>'
    netstat -na | awk ' /^tcp/ { c[$6]++; } END { for (x in c) { print x, c[x]; } }'


    # Postfix mailqueue monitoring
    #
    # Only handle mailq when postfix user is present. The mailq command is also
    # available when postfix is not installed. But it produces different outputs
    # which are not handled by the check at the moment. So try to filter out the
    # systems not using postfix by searching for the postfix user.
    #
    # Cannot take the whole outout. This could produce several MB of agent output
    # on blocking queues.
    # Only handle the last 6 lines (includes the summary line at the bottom and
    # the last message in the queue. The last message is not used at the moment
    # but it could be used to get the timestamp of the last message.
    if which mailq >/dev/null 2>&1 && getent passwd postfix >/dev/null 2>&1; then
    echo '<<<postfix_mailq>>>'
    mailq | tail -n 6
    fi

    #Check status of qmail mailqueue
    if type qmail-qstat >/dev/null
    then
    echo "<<<qmail_stats>>>"
    qmail-qstat
    fi

    # check zpool status
    if [ -x /sbin/zpool ]; then
    echo "<<<zpool_status>>>"
    /sbin/zpool status -x | grep -v "errors: No known data errors"
    fi


    # Statgrab
    # To install: pkg install libstatgrab
    if type statgrab 2>&1 >/dev/null ; then

    statgrab_vars="const. disk. general. page. proc. user."
    statgrab_vars_mem="mem. swap."
    statgrab_sections="proc disk page"

    statgrab $statgrab_vars 1> /tmp/statgrab.$$
    statgrab $statgrab_vars_mem 1>>/tmp/statgrab.$$


    for s in $statgrab_sections
    do
    echo "<<<statgrab_$s>>>"
    grep "^${s}\." /tmp/statgrab.$$ | cut -d. -f2-99 | sed 's/ *= */ /'
    done

    echo '<<<statgrab_net>>>'
    statgrab net. 2>&1 | cut -d. -f2-99 | sed 's/ *= */ /'

    echo '<<<statgrab_mem>>>'
    egrep "^(swap|mem)\." /tmp/statgrab.$$ | sed 's/ *= */ /'

    [ -f /tmp/statgrab.$$ ] && rm -f /tmp/statgrab.$$
    fi


    # Fileinfo-Check: put patterns for files into /etc/check_mk/fileinfo.cfg
    if [ -r "$MK_CONFDIR/fileinfo.cfg" ] ; then
    echo '<<<fileinfo:sep(124)>>>'
    date +%s
    stat -f "%N|%z|%m" $(cat "$MK_CONFDIR/fileinfo.cfg")
    fi


    echo '<<<local>>>'
    if cd $LOCALDIR ; then
    for skript in $(ls) ; do
    if [ -f "$skript" -a -x "$skript" ] ; then
    ./$skript
    fi
    done
    # Call some plugins only every X'th minute
    for skript in [1-9]*/* ; do
    if [ -x "$skript" ] ; then
    run_cached local_${skript//\//\\} ${skript%/*} "$skript"
    fi
    done
    fi


    # Plugins
    if cd $PLUGINSDIR ; then
    for skript in $(ls) ; do
    if [ -f "$skript" -a -x "$skript" ] ; then
    ./$skript
    fi
    done
    # Call some plugins only every X'th minute
    for skript in [1-9]*/* ; do
    if [ -x "$skript" ] ; then
    run_cached plugins_${skript//\//\\} ${skript%/*} "$skript"
    fi
    done
    fi


    # MK's Remote Plugin Executor
    if [ -e "$MK_CONFDIR/mrpe.cfg" ]
    then
    echo '<<<mrpe>>>'
    grep -Ev '^[[:space:]]*($|#)' "$MK_CONFDIR/mrpe.cfg" | \
    while read descr cmdline
    do
    PLUGIN=${cmdline%% *}
    OUTPUT=$(eval "$cmdline")
    echo -n "(${PLUGIN##*/}) $descr $? $OUTPUT" | tr \\n \\1
    echo
    done
    fi

    # 3WARE disk controller (by Radoslaw Bak)
    if type tw_cli 2>&1 >/dev/null ; then
    for C in $(tw_cli show | awk 'NR < 4 { next } { print $1 }'); do
    echo '<<<3ware_info>>>'
    tw_cli /$C show all | egrep 'Model =|Firmware|Serial'
    echo '<<<3ware_disks>>>'
    tw_cli /$C show drivestatus | egrep 'p[0-9]' | sed "s/^/$C\//"
    echo '<<<3ware_units>>>'
    tw_cli /$C show unitstatus | egrep 'u[0-9]' | sed "s/^/$C\//"
    done
    fi