#!/bin/bash # # MOTDs32 powered by Erreur32 # Original Author: Krigler Pavol # Version 0.0.5 # # MOTDs32 is dynamicaly refreshing the /etc/motd file with current informations # about system status and usage. # # Copyright 2013 Pavol Krigler # /etc/motds32/Stats32 > /etc/motds32/Stats32.txt # Path to the configuration file CONFIGFILE="/etc/motds32/motds32.conf" # Message of the day file MOTD='/etc/motd' MOTD32="/etc/motds32/Stats32" PRG32="/etc/motds32" # List of network services which MUST be running NETSERVICE_LIST="/etc/motds32/netservice" # List of processes which MUST be running PROCESS_LIST="/etc/motds32/process" # List of partitions and usage limits PARTITION_TABLE="/etc/motds32/fstab_limits" # List of processes which could be autodiscovered during installation AUTO_PROCESS_DISCOVERY="sshd apache dhcpd named ntpd nscd postfix slapd smbd atd crond smartd dovecot master mysqld" # List of network ports which could be autodiscovered during installation AUTO_NETSERVICE_DISCOVERY="21 22 25 80 443 123 8080" # # Do not modify lines below # VERSION="0.1" LINES=0 TMPDIR="/tmp" STATS_DIR="/var/cache/motds32" COLL_MEM="$TMPDIR/.motds32_mem_$$" COLL_DISK="$TMPDIR/.motds32_disk_$$" COLL_PROC="$TMPDIR/.motds32_proc_$$" COLL_WALL="$TMPDIR/.motdstat_wall_$$" COLL_EMPTY="$TMPDIR/.motds32_empty_$$" # Set default VARIABLES function set_default_values() { DISK_USAGE_WARNING=80 DISK_USAGE_CRITICAL=90 SWAP_USAGE_WARNING=10 SWAP_USAGE_CRITICAL=30 CPU_WARNING=90 CPU_CRITICAL=100 MAX_ROWS_LIMIT=15 MAILQ_WARNING=0 CHECK_NTP="NO" ENABLE_SYSLOG="NO" NTP_PROBLEM_EXEC="" LOAD_TRESHOLD_EXEC="" SWAP_TRESHOLD_EXEC="" DISK_TRESHOLD_EXEC="" } # Clean temporary files test -e $COLL_EMPTY && rm -f $COLL_EMPTY # Make a copy of original MOTD file if exits if [ ! -e "${MOTD}.orig" ]; then test -e ${MOTD} && cp $MOTD ${MOTD}.orig || touch ${MOTD}.orig fi # # Generate stats data # ## not use yet function stats_data() { #make stats file in the good directory $MOTD32 > $PRG32/Stats32.txt } #$stats_data #$MOTD32 > $PRG32/Stats32.txt # # Append message to the report (and to syslog if enabled) # function report() { REPORT=$REPORT"${1}\n" test $ENABLE_SYSLOG = "YES" && logger -t motds32 -- "${1}" } # # Colored text # function red () { echo -e "\e[1;31m$1\e[0;39m" } function green () { echo -e "\e[1;32m$1\e[0;39m" } function yellow () { echo -e "\e[1;33m$1\e[0;39m" } function bold () { echo -e "\033[1m$1\033[0m" } # Trim the line to 24 , (32 is better) characters function normalize() { LENGTH=$(echo $PART | wc -c) if [ $LENGTH -gt 32 ]; then CUT=$(echo $PART | sed -e 's/^\(.......\).*\(.............\)/\1...\2/') PART=$CUT fi } # # Generate list of the most used processess # function gen_process_list() { local PROCESS local PGREP echo "Generating \"process\" configuration file to ${PROCESS_LIST} from autodiscovery" # Generate process file header echo "#" > ${PROCESS_LIST} echo "# List of processes that MUST be running" >> ${PROCESS_LIST} echo "#" >> ${PROCESS_LIST} echo "# (The following processes where found through autodiscovery)" >> ${PROCESS_LIST} test $(which pgrep 2>/dev/null >/dev/null && echo OK || echo NOK) != "OK" && return for PROCESS in ${AUTO_PROCESS_DISCOVERY}; do pgrep -x ${PROCESS} > /dev/null PGREP=$? test ${PGREP} -eq 0 && echo ${PROCESS} >> ${PROCESS_LIST} done } # # Generate netservice list # function gen_netservice_list() { local NETSTAT local ENTRY local RESULT echo "Generating \"netservice\" configuration file to ${NETSERVICE_LIST} from autodiscovery" # Generate process file header echo "#" > ${NETSERVICE_LIST} echo "# List of local network services that MUST be running" >> ${NETSERVICE_LIST} echo "#" >> ${NETSERVICE_LIST} echo "# (The following processes where found through autodiscovery)" >> ${NETSERVICE_LIST} echo "# PROTOCOL IP_ADDRESS:PORT PROCESS NAME" >> ${NETSERVICE_LIST} test $(which netstat 2>/dev/null >/dev/null && echo OK || echo NOK) != "OK" && return NETSTAT=$(netstat -nlp | egrep "^tcp *|^udp *" | sed -e 's/[0-9]*\///' | sed -e 's/LISTEN//' | sed -e 's/: .*$//' | awk '{printf "%-13s%-30s %s\n", $1, $4, $6}') for ENTRY in ${AUTO_NETSERVICE_DISCOVERY}; do RESULT=$(echo "${NETSTAT}" | egrep ":${ENTRY} *") test -n "${RESULT}" && echo "${RESULT}" >> ${NETSERVICE_LIST} done } # Process status function checkprocess { test ! -f ${PROCESS_LIST} && gen_process_list if [ -s ${PROCESS_LIST} ]; then for PROC in $(egrep -v "^#|^$" ${PROCESS_LIST}); do PROCES=$(ps ax | grep "$PROC" | grep -v "grep" | grep -vc "checkclient") if [ $PROCES -gt 0 ]; then if [ $PROCES -eq 1 ]; then green "$PROC" >> $COLL_PROC else #green "$PROC ($PROCES)" | awk '{printf("%-26s %4s\n", $1, $2)}' >> $COLL_PROC green "$PROC ($PROCES)" | awk '{printf("%-27s %4s\n", $1, $2)}' >> $COLL_PROC fi else report "CRITICAL: Process \"${PROC}\" is not running" #red "$PROC is not running" >> $COLL_PROC red "$(printf "%-9.9s is not running" $PROC)" >> $COLL_PROC NOT_RUNNING=1 fi done fi cp $COLL_PROC ${COLL_PROC}.tmp test -n "$NOT_RUNNING" && red " Service status " > $COLL_PROC || green " Service status " > $COLL_PROC cat ${COLL_PROC}.tmp >> $COLL_PROC } # Memory status function memory() { bold "Memory used kB [%]" > $COLL_MEM MEMTOTAL=$(grep "MemTotal" /proc/meminfo | awk '{print $2}') MEMFREE=$(grep "MemFree" /proc/meminfo | awk '{print $2}') SWAPTOTAL=$(grep "SwapTotal" /proc/meminfo | awk '{print $2}') SWAPFREE=$(grep "SwapFree" /proc/meminfo | awk '{print $2}') PERCENTOM=$(echo $MEMFREE $MEMTOTAL | awk '{printf("%d", ($2-$1)/$2*100)}') test $SWAPTOTAL -eq 0 && PERCENTOS=0 || PERCENTOS=$(echo $SWAPFREE $SWAPTOTAL | awk '{printf("%d", ($2-$1)/$2*100)}') SWAP=$(printf "Swap: %13d %3d%%\n" $(($SWAPTOTAL-$SWAPFREE)) $PERCENTOS) # MEMORY USAGE printf "Memory: %11d %3d%%\n" $(($MEMTOTAL-$MEMFREE)) $PERCENTOM >> $COLL_MEM if [ $PERCENTOS -ge $SWAP_USAGE_CRITICAL ] || [ $SWAPTOTAL -eq 0 ]; then red "$SWAP" >> $COLL_MEM CR_MEM=1 report "WARNING: usage of the SWAP space is more than ${SWAP_USAGE_CRITICAL}%" # Execute custom command if [ -n "${SWAP_TRESHOLD_EXEC}" ]; then report "WARNING: executing command: \"$SWAP_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${SWAP_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi else if [ $PERCENTOS -ge $SWAP_USAGE_WARNING ] && [ $PERCENTOS -lt $SWAP_USAGE_CRITICAL ]; then yellow "$SWAP" >> $COLL_MEM WR_MEM=1 report "CRITICAL: usage of the SWAP space is more than ${SWAP_USAGE_WARNING}%" # Execute custom command if [ -n "${SWAP_TRESHOLD_EXEC}" ]; then report "CRITICAL: executing command: \"$SWAP_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${SWAP_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi else echo "$SWAP" >> $COLL_MEM fi fi grep Buffer /proc/meminfo | awk '{printf("%-10s %8s \n",$1,$2)}' >> $COLL_MEM grep "^Cached" /proc/meminfo | awk '{printf("%-10s %8s \n",$1,$2)}' >> $COLL_MEM cp $COLL_MEM ${COLL_MEM}.tmp green " Memory status " > $COLL_MEM test -n "$CR_MEM" && red " Memory status " > $COLL_MEM test -n "$WR_MEM" && yellow " Memory status " > $COLL_MEM cat ${COLL_MEM}.tmp >> $COLL_MEM } # Disk status function disk() { STATUS_DISK=$(egrep -v "^#|^ *#" /etc/motds32/fstab_limits) IFS=$'\n' for DEVICE in $STATUS_DISK do # Read Mount Point path and limits MP=$(echo "${DEVICE}" | awk {'print $1'}) MP_USAGE_WARNING=$(echo "${DEVICE}" | awk {'print $2'}) MP_USAGE_CRITICAL=$(echo "${DEVICE}" | awk {'print $3'}) # If Warning and Critical limit are not set to GLOBAL defaults test -z ${MP_USAGE_WARNING} && MP_USAGE_WARNING=$DISK_USAGE_WARNING test -z ${MP_USAGE_CRITICAL} && MP_USAGE_CRITICAL=$DISK_USAGE_CRITICAL # Fixed problems with long device names and new line PART_RESULT=$(df -h "${MP}" 2>/dev/null) DF_RESULT=$? # Check if the partition exists if [ ${DF_RESULT} -eq 0 ]; then PART=$(echo "${PART_RESULT}" | grep '\/' | tr -d '\n' | awk '{printf("%-13s %5s %4s\n",$6,$4,$5)}' | tr -d '%') normalize USAGE=$(echo $PART | awk '{print $3}') PARTITION=$(echo $PART | awk '{print $1}') if [ $USAGE -ge $MP_USAGE_WARNING ] && [ $USAGE -lt $MP_USAGE_CRITICAL ]; then yellow "${PART}%" >> $COLL_DISK && WR_NOTICE=1 report "WARNING: usage of $PARTITION partition is ${USAGE}% and limit is ${MP_USAGE_WARNING}%" # Execute custom command if [ -n "${DISK_TRESHOLD_EXEC}" ]; then report "WARNING: executing command: \"$DISK_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${DISK_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi continue fi if [ $USAGE -ge $MP_USAGE_CRITICAL ]; then red "${PART}%" >> $COLL_DISK && CR_NOTICE=1 report "CRITICAL: usage of $PARTITION partition is ${USAGE}% and limit is ${MP_USAGE_CRITICAL}%" # Execute custom command if [ -n "${DISK_TRESHOLD_EXEC}" ]; then report "CRITICAL: executing command: \"$DISK_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${DISK_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi fi test $USAGE -lt $MP_USAGE_WARNING && echo "${PART}%" >> $COLL_DISK else red "$(printf "%-13.13s NotMounted" ${MP})" >> $COLL_DISK && CR_NOTICE=1 report "CRITICAL: mount point $MP is not mounted" fi done unset IFS cat $COLL_DISK | sort -k 3 -nr > ${COLL_DISK}.tmp green " Disk status " > $COLL_DISK test -n "$CR_NOTICE" && red " Disk status " > $COLL_DISK test -n "$WR_NOTICE" && yellow " Disk status " > $COLL_DISK bold "partition free usg" >> $COLL_DISK cat ${COLL_DISK}.tmp >> $COLL_DISK } function getmaxrow () { DISK_NUM=$(wc -l $COLL_DISK | awk '{print $1}') test $DISK_NUM -gt $LINES && LINES=$DISK_NUM MEM_NUM=$(wc -l $COLL_MEM | awk '{print $1}') test $MEM_NUM -gt $LINES && LINES=$MEM_NUM PROC_NUM=$(wc -l $COLL_PROC | awk '{print $1}') test $PROC_NUM -gt $LINES && LINES=$PROC_NUM } # Join 3 cols together function join () { getmaxrow MEM_NUM=$(($MEM_NUM + 1)) DISK_NUM=$(($DISK_NUM + 1)) test -e $COLL_EMPTY && rm -f $COLL_EMPTY for I in $(seq $DISK_NUM $LINES); do echo " " >> $COLL_DISK; done for I in $(seq $MEM_NUM $LINES); do echo " " >> $COLL_MEM; done for I in $(seq $LINES); do echo ' | ' >> $COLL_EMPTY; done for I in $(seq $LINES); do echo '|' >> $COLL_WALL; done } # Center text function echo_center { LENGTH=${#1} printf "%$(((80 - $LENGTH) / 2))s" ; echo "$1" } function echo_center2 { LENGTH=${#1} INDENT=`seq 1 $(((78 - $LENGTH) / 2)) | sed 's/.*/ /' | tr -d '\n'` printf "|%-78s|\n" "${INDENT}${1}" } function cat_center { LENGTH=${#1} printf "%$(((80 - $LENGTH) / 2))s" ; cat $1 } # # Check listen services IP:Port # function checkservices { test ! -f ${NETSERVICE_LIST} && gen_netservice_list bold "service(s) (count)" > $COLL_PROC NETSTAT=$(netstat -nlp 2>/dev/null) if [ -e ${NETSERVICE_LIST} ]; then IFS=$'\n' LINE=$(grep -v "^#" ${NETSERVICE_LIST}) for SERVICE in $LINE; do PROTO=$(echo $SERVICE | awk '{print $1}') SOCKET=$(echo $SERVICE | awk '{print $2}') PROCES=$(echo $SERVICE | awk '{print $3}') RESULT=$(echo "$NETSTAT" | egrep "^${PROTO}.*${SOCKET}") # First checking if [ -z "$RESULT" ]; then # DEBUG # echo $NETSTAT >> /tmp/.motds32_netstat_$$.tmp sleep 1 NETSTAT=$(netstat -nlp 2>/dev/null) # Double checking for fake panic RESULT=$(echo "$NETSTAT" | egrep "^${PROTO}.*${SOCKET}") if [ -z "$RESULT" ]; then # red "${PROTO}/${SOCKET}" >> $COLL_PROC red "$(printf "%-24.24s" ${PROTO}/${SOCKET})" >> $COLL_PROC NOT_RUNNING=1 report "CRITICAL: Service \"$PROCES\" not running at socket ${PROTO}/${SOCKET}" else green "$(printf "%-24.24s" ${PROTO}/${SOCKET})" >> $COLL_PROC #green "${PROTO}/${SOCKET}" >> $COLL_PROC fi else #green "${PROTO}/${SOCKET}" >> $COLL_PROC green "$(printf "%-24.24s" ${PROTO}/${SOCKET})" >> $COLL_PROC fi done unset IFS fi } function load { NUM_OF_CPU=$(grep -i processor /proc/cpuinfo | wc -l) LOAD_5MIN=$(awk '{printf $2}' < /proc/loadavg) LOAD_PERCENT=$(echo $LOAD_5MIN $NUM_OF_CPU |awk '{printf "%2.f", $1*100/$2}') LOAD_REPORT="5min load is $LOAD_5MIN on $NUM_OF_CPU cpu(s)" if [ $LOAD_PERCENT -gt $CPU_WARNING ] && [ $LOAD_PERCENT -lt $CPU_CRITICAL ]; then report "WARNING: System load ${LOAD_5MIN} on $NUM_OF_CPU cpu(s)" # Execute custom command if [ -n "${LOAD_TRESHOLD_EXEC}" ]; then report "WARNING: executing command: \"$LOAD_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${LOAD_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi return fi if [ $LOAD_PERCENT -gt $CPU_CRITICAL ]; then report "CRITICAL: system load ${LOAD_5MIN} on $NUM_OF_CPU cpu(s)" # Execute custom command if [ -n "${LOAD_TRESHOLD_EXEC}" ]; then report "CRITICAL: executing command: \"$LOAD_TRESHOLD_EXEC\"" report "--->%---" report "$(eval ${LOAD_TRESHOLD_EXEC} 2>&1)" report "--->%---" fi fi } # # Notify when too much e-mails in mailq # if MAILQ_WARNING=0 skip this test # function check_mailq { test $MAILQ_WARNING -eq 0 && return # Check if mail command is present in $PATH if [ $(which mailq 2>/dev/null >/dev/null && echo OK || echo NOK) != "OK" ]; then echo "WARNING: unable to find mailq command in \$PATH" return fi MAILQ=$(mailq | tail -n 1 | grep -i requests | awk '{print $5}') if [ -n "$MAILQ" ]; then if [ $MAILQ -gt $MAILQ_WARNING ]; then # REPORT=$REPORT"WARNING: mail queue size warning $MAILQ > $MAILQ_WARNING\n" report "WARNING: mail queue size warning $MAILQ > $MAILQ_WARNING" fi fi } # Check NTP status function check_ntp() { test $CHECK_NTP != "YES" && NTPSTATUS=" " && return if [ $(which ntpq 2>/dev/null >/dev/null && echo OK || echo NOK) != "OK" ]; then echo "WARNING: unable to find ntpq command in \$PATH" return fi NTP=$(ntpq -p 2>/dev/null | egrep -v '===|remote|^ ' | sed -e 's/^\(.\).*/\1/' | tr -d '\n') if [ $(echo "$NTP" | grep -c '\*') -eq 0 ]; then test -z "$NTP" && NTPSTATUS=$(red "x") || NTPSTATUS=$(red "$NTP") report "WARNING: NTP synchronization lost" if [ -n "${NTP_PROBLEM_EXEC}" ]; then report "WARNING: executing command: \"$NTP_PROBLEM_EXEC\"" report "--->%---" report "$($NTP_PROBLEM_EXEC 2>&1)" report "--->%---" fi else NTPSTATUS=$(green "$NTP") fi } # # Generate mount point table usage limits from /etc/fstab file # If Warning or Critical mimits are not set user GLOBAL settings # function gen_mount_point_table() { if [ ! -e ${PARTITION_TABLE} ]; then # Exclude list : '#' comments and blacklisted filesystems FSTAB=$(egrep -v '^#|^ .*#|[[:space:]]proc[[:space:]]|[[:space:]]debugfs[[:space:]]|[[:space:]]swap[[:space:]]|[[:space:]]devpts[[:space:]]|[[:space:]]sysfs[[:space:]]|[[:space:]]tmpfs[[:space:]]' < /etc/fstab | awk '{print($2)}') # Generate header echo "# Mount point usage limit Warn [%] Critical [%]" > ${PARTITION_TABLE} for PARTITION in $FSTAB; do printf "%-28s %s %s\n" ${PARTITION} ${DISK_USAGE_WARNING} ${DISK_USAGE_CRITICAL} >> ${PARTITION_TABLE} done fi } # # Main program # # Read the configuration file if [ -e $CONFIGFILE ]; then set_default_values . $CONFIGFILE else echo "Error: configuration file $CONFIGFILE not found" exit 1 fi # display usage function usage { echo "Usage: motds32 OPTIONS" echo " -g, --generate Check system status and generate it to $MOTD file" echo " -s, --status Show limited content of MOTD file" echo " -v, --version Display information about motds32 version and author" echo " -m, --MOTDs32 See MOTD file now!" exit 0 } case $1 in -g|--generate) ;; -s|--status) test -e $MOTD && cat $MOTD exit 0 ;; -v|--version) echo "motds32 version $VERSION" echo "Author: Krigler Pavol, e-mail: motds32@gelogic.net" echo "Custom: Erreur32" exit 0 ;; -m|--MOTDs32) echo "Read motds32 file" cat /etc/motd exit 0 ;; *) usage ;; esac # Generate mount point table usage limits from /etc/fstab file gen_mount_point_table # Create e-mail REPORT header REPORT="System report\n=============\nHost: ${HOSTNAME}\nDate: $(date)\n" # Get disk statistics disk # Get memorty statistics memory # Check local network service checkservices # Check local running processes checkprocess # Join all output to multiple columns join # Check the mailq check_mailq # Check NTP status check_ntp # Create MOTD system status header load echo_center "${HOSTNAME}${NTPSTATUS} > status at $(date +"%R") > ${LOAD_REPORT}" bold > $MOTD echo >> $MOTD # cat /etc/motds32/Stats32.txt >> $MOTD paste -d "" $COLL_DISK $COLL_EMPTY $COLL_MEM $COLL_EMPTY $COLL_PROC | head -n $MAX_ROWS_LIMIT >> $MOTD #paste -d "" $COLL_DISK $COLL_EMPTY $COLL_MEM $COLL_EMPTY $COLL_PROC >> $MOTD cat /etc/motds32/Stats32.txt >> $MOTD # Full report #paste -d "" $COLL_DISK $COLL_EMPTY $COLL_MEM $COLL_EMPTY $COLL_PROC > $MOTD.full echo >> $MOTD # Append original MOTD from /etc/motd.orig file if exists if [ -s ${MOTD}.orig ]; then # echo_center "--- Message Of The Day --- (from /etc/motd.orig file) ---" >> $MOTD # echo >> $MOTD cat ${MOTD}.orig >> $MOTD fi # Create statistics directory test ! -e $STATS_DIR && mkdir -p $STATS_DIR # Check if server has rebooted (current uptime < stored uptime if [ -s $STATS_DIR/uptime ]; then if [ $(cut -d '.' -f 1 < /proc/uptime) -lt $(cat $STATS_DIR/uptime) ]; then report "CRITICAL: System has rebooted." fi fi # Update uptime statistics test -e /proc/uptime && cut -d '.' -f 1 < /proc/uptime > $STATS_DIR/uptime # Removing temporary files rm -f ${COLL_PROC} ${COLL_PROC}.tmp rm -f ${COLL_MEM} ${COLL_MEM}.tmp rm -f ${COLL_DISK} ${COLL_DISK}.tmp rm -f ${COLL_EMPTY} rm -f ${COLL_WALL} if [ ! -z "$EMAIL" ]; then # Check if mail command is present in $PATH if [ $(which mail 2>/dev/null >/dev/null && echo OK || echo NOK) != "OK" ]; then echo "WARNING: unable to find mail command in \$PATH" else if [ ! -z "$(echo -e "$REPORT" | egrep -i "warning:|critical:")" ]; then echo -e "$REPORT" | mail "$EMAIL" -s "MOTDs32 alert" fi fi fi