check_cpu

#!/bin/bash
# ======================================================================
#
# Check CPU usage
#
# requirements:
# - top
# - bc
#
# ----------------------------------------------------------------------
# 2020-03-10  v1.0  <axel.hahn@iml.unibe.ch>
# 2020-03-23  v1.1  <axel.hahn@iml.unibe.ch> added more data
# 2020-07-08  v1.2  <axel.hahn@iml.unibe.ch> FIX: set "ph." instead "ps."
# 2020-07-17  v1.3  <axel.hahn@iml.unibe.ch> use ph.require to check binaries
# ======================================================================


. `dirname $0`/inc_pluginfunctions
tmpfile=/tmp/check_cpu_$$

# ----------------------------------------------------------------------
# MAIN
# ----------------------------------------------------------------------

# --- check required tools
ph.require bc top


# --- check param -h
if [ "$1" = "-h" ]; then
    echo "
    usage: $0 [ -w value -c value -h ]

        -w  Warning level
        -c  Critical level
        -h  this help
    "
    exit 0
fi

# set default / override from command line params
typeset -i iWarnLimit=`     ph.getValueWithParam 75 w "$@"`
typeset -i iCriticalLimit=` ph.getValueWithParam 90 c "$@"`

# get cpu status i.e.
# %Cpu(s): 33.3 us,  9.5 sy,  0.0 ni, 57.1 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
#   us, user    : time running un-niced user processes
#   sy, system  : time running kernel processes
#   ni, nice    : time running niced user processes
#   id, idle    : time spent in the kernel idle handler
#   wa, IO-wait : time waiting for I/O completion
#   hi : time spent servicing hardware interrupts
#   si : time spent servicing software interrupts
#   st : time stolen from this vm by the hypervisor
top -b -n 1 | head -5 | grep "^\%Cpu" >$tmpfile

cpuUser=`   awk '{ print $2 }' $tmpfile`
cpuSystem=` awk '{ print $4 }' $tmpfile`
cpuNice=`   awk '{ print $6 }' $tmpfile`
cpuIdle=`   awk '{ print $8 }' $tmpfile`
cpuWait=`   awk '{ print $10 }' $tmpfile`
cpuHi=`     awk '{ print $12 }' $tmpfile`
cpuSi=`     awk '{ print $14 }' $tmpfile`
cpuSt=`     awk '{ print $16 }' $tmpfile`
cpuNonIdle=`echo 100-$cpuIdle | bc`

rm -f $tmpfile

if [ "$(echo "${cpuNonIdle} > ${iWarnLimit}" | bc)" -eq 1 ]; then

    if [ "$(echo "${cpuNonIdle} > ${iCriticalLimit}" | bc)" -eq 1 ]; then
        ph.setStatus "critical"
    else
        ph.setStatus "warning"
    fi
fi

# --- status output
ph.status "CPU-USAGE [%] ${cpuNonIdle} ... user: ${cpuUser} - system: ${cpuSystem} - idle: ${cpuIdle}"


echo "

Legend:

   hwi - Time spent handling hardware interrupt routines. (Whenever a peripheral unit want attention form the CPU, it literally pulls a line, to signal the CPU to service it)
   swi - Time spent handling software interrupt routines. (a piece of code, calls an interrupt routine...)
    st - Time spent on involuntary waits by virtual cpu while hypervisor is servicing another processor (stolen from a virtual machine)

  nice - Time spent running niced user processes (User defined priority)
  wait - Time spent on waiting on IO peripherals (eg. disk)

system - Time spent in kernel space
  user - Time spent in user space
  idle - Time spent in idle operations
"


# --- performance data usage
ph.perfadd "cpu-usage"    "${cpuNonIdle}" $iWarnLimit $iCriticalLimit 0 100

# for graphite module: send limits
# ph.perfadd "cpu-warn"     $iWarnLimit      "" "" 0 100
# ph.perfadd "cpu-crit"     $iCriticalLimit  "" "" 0 100

# --- performance data single values
ph.perfadd "cpu-system"   "${cpuSystem}"   "" "" 0 100
ph.perfadd "cpu-user"     "${cpuUser}"     "" "" 0 100
ph.perfadd "cpu-idle"     "${cpuIdle}"     "" "" 0 100
ph.perfadd "cpu-nice"     "${cpuNice}"     "" "" 0 100
ph.perfadd "cpu-wait"     "${cpuWait}"     "" "" 0 100
ph.perfadd "cpu-hwi"      "${cpuHi}"       "" "" 0 100
ph.perfadd "cpu-swi"      "${cpuSi}"       "" "" 0 100
ph.perfadd "cpu-st"       "${cpuSt}"       "" "" 0 100

ph.exit

# ----------------------------------------------------------------------