get rid of some bashisms / clean up. add ping metric. add --metrics cli option

README.md

@@ -1,3 +1,52 @@
+# metrics.sh
+
+metrics.sh is a metrics collection and fowarding daemon implemented in portable shell scripts. A simple interface based on hooks allows for writing custom metrics collectors.
+
+## Usage
+
+```
+  Usage: ./metrics.sh [-d] [-h] [-v] [-r reporter] [-i interval]
+
+  Options:
+
+    -r, --reporter <reporter>  use specified reporter (default: stdout)
+    -i, --interval <seconds>   collect metrics every n seconds (default: 2)
+    -v, --verbose              enable verbose mode
+    -d, --docs                 show documentation
+    -h, --help                 show this text
+```
+
+## Installation
+
+```bash
+$ git clone git@github.com:pstadler/metrics.sh.git
+```
+
+TODO: /etc/init.d
+
+### Requirements
+
+metrics.sh has been tested on Ubuntu and Mac OS X but is supposed to run on most *NIX-line operating systems. Some of the provided metrics require [procfs](http://en.wikipedia.org/wiki/Procfs) to be available (check with `[ -d /proc ] && echo available || echo not available`).
+
+## Metrics
+
+Metric        | Description
+------------- | -------------
+cpu           | CPU usage in %
+memory        | Memory usage in %
+swap          | Swap usage in %
+network_io    | Network I/O in kB/s
+disk_io       | Disk I/O in MB/s
+disk_usage    | Disk usage in %
+heartbeat     | System heartbeat
+ping          | Check if remote host is reachable
+
+TODO: how to write custom metrics
+
+## Reporters
+
+TODO: how to write custom reporters
+
 ## TODO
 
 - README

lib/main.sh

@@ -1,9 +1,13 @@
+#!/bin/sh
+
 # load utils
-for util in ./lib/utils/*.sh; do source $util; done
+for util in ./lib/utils/*.sh; do
+  . $util
+done
 
 # init
-__METRICS=()
-__REPORTERS=()
+__AVAILABLE_METRICS=
+__AVAILABLE_REPORTERS=
 
 main_load () {
   # load reporter
@@ -12,14 +16,14 @@ main_load () {
     local reporter=${filename%.*}
 
     # source reporter and copy functions
-    source $file
+    . $file
     copy_function init      __r_${reporter}_init
     copy_function report    __r_${reporter}_report
     copy_function terminate __r_${reporter}_terminate
     copy_function docs      __r_${reporter}_docs
     unset -f init report terminate docs
 
-    __REPORTERS+=($reporter)
+    __AVAILABLE_REPORTERS=$(trim "$__AVAILABLE_REPORTERS $reporter")
   done
 
   # load metrics
@@ -28,7 +32,7 @@ main_load () {
     local metric=${filename%.*}
 
     # soruce metric and copy functions
-    source $file
+    . $file
     copy_function init      __m_${metric}_init
     copy_function collect   __m_${metric}_collect
     copy_function terminate __m_${metric}_terminate
@@ -36,11 +40,16 @@ main_load () {
     unset -f init collect terminate docs
 
     # register metric
-    __METRICS+=($metric)
+    __AVAILABLE_METRICS=$(trim "$__AVAILABLE_METRICS $metric")
   done
 }
 
 main_init () {
+  # handle args
+  __METRICS=$(echo $1 | sed 's/,/ /g')
+  __REPORTER=$2
+
+  # create temp dir
   TEMP_DIR=$(make_temp_dir)
 
   # register trap
@@ -49,13 +58,27 @@ main_init () {
     trap - SIGTERM && kill -- -$$ SIGINT SIGTERM EXIT
   ' SIGINT SIGTERM EXIT
 
+  # check if reporter exists
+  if ! in_array $__REPORTER "$__AVAILABLE_REPORTERS"; then
+    echo "Error: reporter '$__REPORTER' is not available"
+    exit 1
+  fi
+
+  # check if metrics exist
+  for metric in $__METRICS; do
+    if ! in_array $metric "$__AVAILABLE_METRICS"; then
+      echo "Error: metric '$metric' is not available"
+      exit 1
+    fi
+  done
+
   # init reporter
-  if is_function __r_${REPORTER}_init; then
-    __r_${REPORTER}_init
+  if is_function __r_${__REPORTER}_init; then
+    __r_${__REPORTER}_init
   fi
 
   # init metrics
-  for metric in ${__METRICS[@]}; do
+  for metric in $__METRICS; do
     if ! is_function __m_${metric}_init; then
       continue
     fi
@@ -76,32 +99,33 @@ main_collect () {
       _r_result="$2"
     fi
     if is_number $_r_result; then
-      __r_${REPORTER}_report $_r_label $_r_result
+      __r_${__REPORTER}_report $_r_label $_r_result
     fi
   }
 
   # collect metrics
-  for metric in ${__METRICS[@]}; do
+  for metric in $__METRICS; do
     if ! is_function __m_${metric}_collect; then
       continue
     fi
 
     fork () {
-      __m_${metric}_collect
-      sleep $INTERVAL
-      fork
+      while true; do
+        __m_${metric}_collect
+        sleep $INTERVAL
+      done
     }
     fork &
     unset -f fork
   done
 
   # run forever
-  tail -f /dev/null # `sleep infinity` is not portable
+  sleep 2147483647 # `sleep infinity` is not portable
 }
 
 main_terminate () {
   # terminate metrics
-  for metric in ${__METRICS[@]}; do
+  for metric in $__METRICS; do
     if ! is_function __m_${metric}_terminate; then
       continue
     fi
@@ -109,8 +133,8 @@ main_terminate () {
   done
 
   # terminate reporter
-  if is_function __r_${REPORTER}_terminate; then
-    __r_${REPORTER}_terminate
+  if is_function __r_${__REPORTER}_terminate; then
+    __r_${__REPORTER}_terminate
   fi
 
   # delete temporary directory
@@ -121,7 +145,7 @@ main_terminate () {
 
 main_docs () {
   echo "# Metrics"
-  for metric in ${__METRICS[@]}; do
+  for metric in $__AVAILABLE_METRICS; do
     if ! is_function __m_${metric}_docs; then
       continue
     fi
@@ -133,7 +157,7 @@ main_docs () {
 
   echo
   echo "# REPORTERS"
-  for reporter in ${__REPORTERS[@]}; do
+  for reporter in $__AVAILABLE_REPORTERS; do
     if ! is_function __r_${reporter}_docs; then
       continue
     fi

lib/utils/functions.sh

@@ -1,11 +1,11 @@
 #!/bin/sh
 
 is_function () {
-  [ "`type -t $1`" == 'function' ]
+  declare -f -F $1 > /dev/null; return $?
 }
 
 # http://stackoverflow.com/a/1369211/183097
 copy_function () {
-  declare -F $1 > /dev/null || return 1
+  is_function $1 || return 1
   eval "$(echo "${2}()"; declare -f ${1} | tail -n +2)"
 }

lib/utils/helpers.sh

@@ -1,7 +1,19 @@
+#!/bin/sh
+
 is_number () {
-  [ ! -z "$1" ] && printf '%f' "$1" &>/dev/null
+  [ ! -z "$1" ] && printf '%f' "$1" > /dev/null 2>&1
 }
 
 iso_date () {
   date -u +"%Y-%m-%dT%H:%M:%SZ"
+}
+
+in_array () {
+  local item=$1
+  local arr=$2
+  [[ " $2 " = *" $1 "* ]]
+}
+
+trim () {
+  echo $1 | sed -e 's/^[[:space:]]*//g' -e 's/[[:space:]]*\$//g'
 }

lib/utils/os.sh

@@ -1,6 +1,6 @@
 #!/bin/sh
 
-declare -r OS_TYPE=$(case "$OSTYPE" in
+readonly OS_TYPE=$(case "$OSTYPE" in
   (solaris*) echo solaris;;
   (darwin*)  echo osx;;
   (linux*)   echo linux;;
@@ -8,11 +8,11 @@ declare -r OS_TYPE=$(case "$OSTYPE" in
   (*)        echo unknown;;
 esac)
 
-is_solaris () { [ $OS_TYPE == 'solaris' ]; }
-is_osx ()     { [ $OS_TYPE == 'osx' ]; }
-is_linux ()   { [ $OS_TYPE == 'solaris' ]; }
-is_bsd ()     { [ $OS_TYPE == 'bsd']; }
-is_unknown () { [ $OS_TYPE == 'unknown' ]; }
+is_solaris () { [ $OS_TYPE = 'solaris' ]; }
+is_osx ()     { [ $OS_TYPE = 'osx' ]; }
+is_linux ()   { [ $OS_TYPE = 'solaris' ]; }
+is_bsd ()     { [ $OS_TYPE = 'bsd']; }
+is_unknown () { [ $OS_TYPE = 'unknown' ]; }
 
 # http://unix.stackexchange.com/a/84980/50905
 if is_osx; then

lib/utils/verbose.sh

@@ -1,3 +1,5 @@
+#!/bin/sh
+
 VERBOSE_MODE=false
 
 verbose_on () {

metrics.sh

@@ -3,6 +3,7 @@
 # config
 INTERVAL=2
 REPORTER=stdout
+METRICS=cpu,disk_io,disk_usage,heartbeat,memory,network_io,swap
 
 # env
 LC_ALL=en_US.UTF-8
@@ -15,7 +16,7 @@ opt_docs=false
 opt_verbose=false
 
 usage () {
-  echo "  Usage: $0 [-d] [-h] [-v] [-r reporter] [-i interval]"
+  echo "  Usage: $0 [-d] [-h] [-v] [-m metrics] [-r reporter] [-i interval]"
 }
 
 help () {
@@ -24,16 +25,22 @@ help () {
   echo
   echo "  Options: "
   echo
-  echo "    -r, --reporter <reporter>  use specified reporter (default: stdout)"
-  echo "    -i, --interval <seconds>   collect metrics every n seconds (default: 2)"
-  echo "    -v, --verbose              enable verbose mode"
-  echo "    -d, --docs                 show documentation"
-  echo "    -h, --help                 show this text"
+  echo "    -m, --metrics  <metric1,...>  use specified metrics"
+  echo "    -r, --reporter <reporter>     use specified reporter (default: stdout)"
+  echo "    -i, --interval <seconds>      collect metrics every n seconds (default: 2)"
+  echo "    -v, --verbose                 enable verbose mode"
+  echo "    -d, --docs                    show documentation"
+  echo "    -h, --help                    show this text"
   echo
 }
 
 while [ $# -gt 0 ]; do
   case $1 in
+    -m|--metrics)
+      shift
+      METRICS=$1
+      ;;
+
     -r|--reporter)
       shift
       REPORTER=$1
@@ -66,9 +73,8 @@ while [ $# -gt 0 ]; do
   shift
 done
 
-
 # run
-source ./lib/main.sh
+. ./lib/main.sh
 
 if [ $opt_verbose = "true" ]; then
   verbose_on
@@ -77,16 +83,17 @@ fi
 verbose "OS detected: $OS_TYPE"
 
 main_load
-verbose "Metrics loaded: ${__METRICS[@]}"
-verbose "Reporters loaded: ${REPORTER}"
+verbose "Available metrics: $__AVAILABLE_METRICS"
+verbose "Available reporters: $__AVAILABLE_REPORTERS"
 
 if [ "$opt_docs" = true ]; then
   main_docs
   exit
 fi
 
-main_init
-verbose "Metrics initialized"
+main_init $METRICS $REPORTER
+verbose "Using metrics: $__METRICS"
+verbose "Using reporter: $__REPORTER"
 
 verbose "Collecting metrics every $INTERVAL second(s)"
 main_collect

metrics/heartbeat.sh

@@ -0,0 +1,9 @@
+#!/bin/sh
+
+collect () {
+  report 1
+}
+
+docs () {
+  echo "Send a simple heartbeat in form of an integer '1'."
+}

metrics/network_io.sh

@@ -8,7 +8,7 @@ init () {
       NETWORK_IO_INTERFACE="eth0"
     fi
   fi
-  readonly __network_io_divisor=$[$INTERVAL * 1024]
+  readonly __network_io_divisor=$(($INTERVAL * 1024))
 }
 
 if is_osx; then

metrics/ping.sh

@@ -1,9 +1,23 @@
 #!/bin/sh
 
+init () {
+  if [ -z $PING_REMOTE_HOST ]; then
+    echo "Error: ping metric requires \$PING_REMOTE_HOST to be specified"
+    exit 1
+  fi
+}
+
 collect () {
-  report 1
+  ping -c 1 $PING_REMOTE_HOST > /dev/null 2>&1
+  if [ $? -eq 0 ]; then
+    report 1
+  else
+    report 0
+  fi
 }
 
 docs () {
-  echo "Send a simple ping in form of an integer '1'."
+  echo "Check if remote host is reachable by sending a single ping."
+  echo "Reports '1' if ping was successful, '0' if not."
+  echo "\$PING_REMOTE_HOST="
 }

reporters/influxdb.sh

@@ -12,6 +12,7 @@ init () {
 
   if [ "$INFLUXDB_SEND_HOSTNAME" = true ]; then
     __influxdb_columns="[\"value\",\"host\"]"
+    __influxdb_hostname=$(hostname)
   else
     __influxdb_columns="[\"value\"]"
   fi
@@ -22,7 +23,7 @@ report () {
   local value=$2
   local points
   if [ "$INFLUXDB_SEND_HOSTNAME" = true ]; then
-    points="[$value,\"$HOSTNAME\"]"
+    points="[$value,\"$__influxdb_hostname\"]"
   else
     points="[$value]"
   fi

reporters/keen_io.sh

@@ -12,7 +12,7 @@ init() {
   fi
 
   if [ -z $KEEN_IO_EVENT_COLLECTION ]; then
-    KEEN_IO_EVENT_COLLECTION=$HOSTNAME
+    KEEN_IO_EVENT_COLLECTION=$(hostname)
   fi
 
   __keen_io_api_url="https://api.keen.io/3.0"
@@ -33,5 +33,5 @@ docs () {
   echo "Send data to Keen IO (https://keen.io)."
   echo "\$KEEN_IO_WRITE_KEY=<write_key>"
   echo "\$KEEN_IO_PROJECT_ID=<project_id>"
-  echo "\$KEEN_IO_EVENT_COLLECTION=$HOSTNAME"
+  echo "\$KEEN_IO_EVENT_COLLECTION=$(hostname)"
 }