check_docker.sh

#!/bin/bash

# Default threshold values.
cpu_threshold_warning=25
cpu_threshold_critical=50
mem_threshold_warning=50
mem_threshold_critical=75
filter=''

print_help() {
    echo "usage: ${0} [OPTIONS]"
    echo "options:"
    echo -e " -c, --cpu <warning:critical>\t\tthreshold values for cpu usage"
    echo -e " -f, --filter <regex>\t\t\tcontainers to check"
    echo -e " -h, --help\t\t\t\tprint this help"
    echo -e " -m, --memory <warning:critical>\tthreshold values for memory usage"
}

while [[ $# -gt 0 ]]; do
    case "$1" in
        -c|--cpu)
            cpu_threshold_warning=$(/usr/bin/awk -F ':' '{print +$1}' <(echo $2))
            cpu_threshold_critical=$(/usr/bin/awk -F ':' '{print +$2}' <(echo $2))
            shift 2
            ;;

        -f|--filter)
            filter=$2
            shift 2
            ;;

        -m|--memory)
            mem_threshold_warning=$(/usr/bin/awk -F ':' '{print +$1}' <(echo $2))
            mem_threshold_critical=$(/usr/bin/awk -F ':' '{print +$2}' <(echo $2))
            shift 2
            ;;

        --help|-h)
            print_help
            exit 0
            ;;

        *)
            echo "Error while parsing argument '$1'."
            exit 1
            ;;
    esac
done

# Gathering facts.
all_containers=$(sudo /usr/bin/docker ps --format '{{.Names}}' --filter "name=${filter}" --no-trunc --quiet --all)
running_containers=$(sudo /usr/bin/docker ps --format '{{.Names}}' --filter "name=${filter}" --no-trunc --quiet)
running_containers_stats=$(sudo /usr/bin/docker stats --format '{{.Name}}:{{.CPUPerc}}:{{.MemPerc}}' --filter "name=${filter}" --no-stream)

not_running_containers=$(comm -23 <(echo $all_containers | /usr/bin/tr ' ' '\n') <(echo $running_containers | /usr/bin/tr ' ' '\n'))

# Format data.
output=''
perf=''

cpu_scale=";${cpu_threshold_warning};${cpu_threshold_critical};0;100"
mem_scale=";${mem_threshold_warning};${mem_threshold_critical};0;100"

# Running containers.
for container in ${running_containers}; do
    output+="OK: ${container} status is running\n"
done
# Not running containers.
for container in ${not_running_containers}; do
    output+="NOK: ${container} status is not running\n"
done

# CPU and memory statistics for running containers.
for container_stats in ${running_containers_stats}; do
    name=$(/usr/bin/awk -F ':' '{print $1}' <(echo $container_stats))
    [[ ! $name =~ ${filter} ]] && continue
    all_cpu=$(/usr/bin/awk -F ':' '{print +$2}' <(echo $container_stats))
    mem=$(/usr/bin/awk -F ':' '{print +$3}' <(echo $container_stats))
    proc_number=$(nproc --all)
    cpu=$(echo "${all_cpu} ${proc_number}" | /usr/bin/awk '{printf "%.2f", $1 / $2}')

    status='OK'
    [[ ${cpu%.*} -ge $cpu_threshold_warning ]] && status='WARNING'
    [[ ${cpu%.*} -ge $cpu_threshold_critical ]] && status='CRITICAL'
    output+="${status}: ${name} cpu is ${cpu}%\n"

    status='OK'
    [[ ${mem%.*} -ge $mem_threshold_warning ]] && status='WARNING'
    [[ ${mem%.*} -ge $mem_threshold_critical ]] && status='CRITICAL'
    output+="${status}: ${name} memory is ${mem}%\n"

    perf+="${name}_cpu=${cpu}%${cpu_scale} "
    perf+="${name}_mem=${mem}%${mem_scale} "
done

# Nagios OK status.
exit_code=0
# Nagios WARNING status.
grep --quiet 'WARNING:' <(echo $output) && exit_code=1
# Nagios CRITICAL status.
grep --quiet 'NOK:\|CRITICAL:' <(echo $output) && exit_code=2

# Plugin output.
echo "${output%'\n'}|${perf%' '}"
exit $exit_code