Dotaz: Nagios a Centreon

[Dalibor V.]

Mam 3 problemy s Centreonem (resp. Nagiosem) a prosim Vas o pomoc:

 

- v konfiguraci služeb mi chybí pole ARGS pro pridani promennych parametru k prikazu

- nevykresluji se grafy (resp. necitelny png obrazek)

- a nedari se mi spustit command, ktery potrebuje prava roota (prostrednictvim sudo). Neznam spravnou syntaxi, z konzole OK

 

Diky za pomoc. Predpokladam, ze nejsem jediny, co pouziva Nagios v kombinaci s Centreonem.

Odpovědi

[komodo]

1. http://forge.centreon.com/issues/2479

2. jak vypada obrazek ? hlasi nejakou chybu v logu ?

3. co znamena nedari se spustit ? jak vypada definice commandu a nasledna definice checku ?

[skunerq]

ad 2, v nastaveni chybi cesta k rrd souborum, nevi kde je rrdtools ... zapnete v centreonovi kompletni logovani {administration=>options=>debug} a v /usr/local/centreon/log/ zkouknete soubory centreonstorage.log a rrdtool.log Dale zkontrolujte spravne nastavene cesty v centreonstorage {administration=>options=>centreonstorage=>options}

ad 3, v /etc/sudoers by melo byt neco takoveho

## BEGIN: CENTREON SUDO #Add by CENTREON installation script User_Alias CENTREON=www-data,nagios Defaults:CENTREON !requiretty # Nagios Restart CENTREON ALL = NOPASSWD: /etc/init.d/nagios* restart CENTREON ALL = NOPASSWD: /etc/init.d/nagios restart # Nagios reload CENTREON ALL = NOPASSWD: /etc/init.d/nagios* reload CENTREON ALL = NOPASSWD: /etc/init.d/nagios reload # Nagios test config CENTREON ALL = NOPASSWD: /usr/local/nagios/bin/nagios* -v * CENTREON ALL = NOPASSWD: /usr/local/nagios/bin/nagios -v * # Nagios test for optim config CENTREON ALL = NOPASSWD: /usr/local/nagios/bin/nagios* -s * CENTREON ALL = NOPASSWD: /usr/local/nagios/bin/nagios -s * # Snmptrapd Restart CENTREON ALL = NOPASSWD: /etc/init.d/snmptrapd restart # CentStorage Restart CENTREON ALL = NOPASSWD: /etc/init.d/centstorage restart # CentStorage stop CENTREON ALL = NOPASSWD: /etc/init.d/centstorage stop # CentStorage start CENTREON ALL = NOPASSWD: /etc/init.d/centstorage start ## END: CENTREON SUDO

[Marek]

Ja mam asi podobny problem. V logach zadne chyby, cesty nastaveny spravne.. Pokud chcete, poslu nejake informace na ukazku, treba neco vyctete, ale ja zadne chyby nevidim

[skunerq]

jsou ve /var/lib/centreon, adresare metrics a status? jsou v techto adresarich .rrd soubory? vykresluje centreon preformance grafy? {home=>nagios statistics=>graphs}

zkuste zapnout debug v centreonstorage {/usr/local/centreon/bin/centreonstorage} v defaultu je vypnuto {$debug = 1;}. pak restartnete centreonstorage servisu nasledne zkouknete /usr/local/centreon/log/centreonstorage.log/

[Marek]

ano - ano - NE - ok

[Vasser]

Příloha:

• centreon.jpg (46733 bytů)

rrd se vytvari:

19/7/2011 10:33:59 - DB updating : /var/lib/centreon/metrics/234.rrd
19/7/2011 10:33:59 - DB updating : /var/lib/centreon/status/119.rrd
19/7/2011 10:33:59 - DB updating : /var/lib/centreon/metrics/226.rrd
19/7/2011 10:33:59 - DB updating : /var/lib/centreon/metrics/227.rrd
19/7/2011 10:33:59 - DB updating : /var/lib/centreon/status/115.rrd

ale vypada to takto.. viz. priloha

[Ruža Becelin]

Predpokladam, ze php-gd balik je nainstalovany?

[Vasser]

neni

apt-get install php-gd
Čtu seznamy balíků… Hotovo
Vytvářím strom závislostí
Čtu stavové informace… Hotovo
E: Nelze najít balík php-gd

k cemu to je?

[Vasser]

toto ale mam:

apt-get install php5-gd
Čtu seznamy balíků… Hotovo
Vytvářím strom závislostí
Čtu stavové informace… Hotovo
php5-gd je již nejnovější verze.

[skunerq]

tak uz zustava perf-data v kongiguraci nagios.cfg zkontrolujte tyto polozky

process_performance_data=1
service_perfdata_command=process-service-perfdata
service_perfdata_file=/usr/local/nagios/var/service-perfdata
service_perfdata_file_mode=2
event_broker_options=-1
broker_module=/usr/local/nagios/bin/ndomod-3x.o config_file=/usr/local/nagios/etc/ndomod.cfg

a zkontroloval zda ma webserver opravneni do slozek /var/lib/centreon/

[Vasser]

Zdravim a dekuju za reakci. Vse mam nastaveno (postupoval jsem podle navodu), ale bohuzel problem pretrvava,. Jsem bezradnej

[skunerq]

mozna by bylo lepsi postupovat podle oficialnich navodu link

[komodo]

Nebo to jeste projit dle toho.

[Vasser]

To uz jsem taky procital, ale mam "vsechno" nastaveno dobre. Netusim, kde je zakopanej pes.

[Dalibor V.]

Koukam ze se to tady docela rozjelo

skript

$USER1$/check_asterisk_peers -p 100

poustim pres centreon v Configuration > Commands > Checks a vysledek je takovy:

Plugin test
Command Line	/usr/lib/nagios/plugins/check_asterisk_peers -p 100
Output	CRITICAL: ?[0;37mUnable to connect to remote asterisk (does /var/run/asterisk/asterisk.ctl exist?)
?[0m
Status	CRITICAL

ALE, kdyz ho pustim v terminalu pod uzivatelem nagios (pod tim bezi i centreon) se sudo (v /etc/sudoers mam potrebne udaje, tudiz na tento prikaz nepotrebuje uzivatel nagios heslo), vysledek je ok:

server:~# su nagios
server:/$ whoami
nagios
server:/$ sudo /usr/lib/nagios/plugins/check_asterisk_peers -p 100
 OK: 100

Kam do toho commandu v centreonu mam zadat sudo?

[skunerq]

jakeho ma adresar plugin vlastnika ?

ja mam pluginy v /usr/loval/nagios/libexec a vlastnikem je nagios

drwxrwsr-x  3 nagios nagios   20480 2011-07-15 18:52 libexec

[Dalibor V.]

drwxr-xr-x 3 root root 12288 20. čec 12.27 plugins

Takze to precte kazdy. Oni ty pluginy jdou spoustet, to je ok. Ale kdyz uz ten samotny plugin potrebuje prava roota, tak ho musim pustit pod sudo (/etc/sudoers mi resi to, ze neni zapotrebi vyzva k zadani hesla korektne). Ale to nevim, jak to z toho centreonu udelat, protoze tam pred command sudo nejde priradit.

[Ruža Becelin]

Co pouzit alias?

[skunerq]

podle me bys stacilo

chown nagios:nagios /usr/lib/nagios/plugins/*

[Ruža Becelin]

To ne, protoze sudo potrebuji pluginy...

[skunerq]

jediny kdo potrebuje ty pluginy je nagios, a root si je pusti taky bez problemu. Centreon to ma osetreno pres sudoers, takze nevidim duvod

[Ruža Becelin]

Divam jsem se na moji konfiguraci a pro spusteni binarky v privilegovanem modu mam sudo primo v pluginu.

Takze:
1) Bud si udela alias a sudo bude poustet pres nej
2) Spusti sudo primo v pluginu

[Dalibor V.]

To jakoze napisu do pluginu nekam na zacatek sudo?

[Dalibor V.]

tady je ten script, kam mam dat sudo?

#!/bin/sh

# Check asterisk peers plugin for Nagios.
# Written by Chad Phillips (chad@apartmentlines.com)
# Last Modified: 2009-05-02

ASTERISK=/usr/sbin/asterisk

PROGPATH=`dirname $0`
REVISION=`echo '$Revision: 2 $' | sed -e 's/[^0-9.]//g'`

. $PROGPATH/utils.sh

print_usage() {
    echo "
Usage: check_asterisk_peers [--type | -t sip|iax] [--peers | -p peers] [--registrations | -r registrations] [--verify-peers] [--verify-registrations] [--config-file filepath]
Usage: check_asterisk_peers --help | -h

Description:

This plugin checks Asterisk peers via Asterisk CLI commands to make sure
they are in a healthy state.  Unlike other monitoring plugins, status
is obtained from the perspective of the Asterisk server -- it's a good plugin
to use for monitoring the state of your connections to providers.

Both the peer itself and registration status to a peer can be monitored,
with the option of verifying that the peer/registration is actively configured.

If any of the checked peers/registrations return a non-OK state, a critical
state is returned. If verification is being used and a peer/registration is
not verified as being in the active configuration, then the status check is
skipped for that item instead of a critical state being returned.

Tested to work on Linux.

The following arguments are accepted:

  --type  | -t            (Optional) The type of peer to check.  Valid values
                          are sip, iax. Defaults to sip.

  --peers | -p            (Optional) A space separated list of peers to check.
                          Use the peer name given in the configuration,
                          without the surrounding brackets.

  --registrations | -r    (Optional) A space separated list of registrations
                          to check, in the form of username@uri. For example,
                          to check the registration of peer foo to
                          somesite.com, use foo@somesite.com

  --verify-peers          (Optional) If set, peer entries are verified in
                          the configuration, and the check is skipped
                          if they are not found.

  --verify-registrations  (Optional) If set, registrations are verified in
                          the configuration, and the check is skipped
                          if they are not found.

  --config-file           (Optional) The location of the configuration file(s).
                          Use a space separated list if specifying multiple
                          files.

                          Default value is based on the --type setting:

                          sip: /etc/asterisk/sip.conf
                          iax: /etc/asterisk/iax.conf

  --help | -h             Print this help and exit.

Examples:

Check peers [foo] and [bar]:

  check_asterisk_peers -p \"foo bar\"

Check peers [foo] and [bar], and registration for usr baz at somesite.com,
and verify that both the peers and the registration are actively configured.

  check_asterisk_peers -p \"foo bar\" -r \"baz@somesite.com\" --verify-peers --verify-registrations

Caveats:

The verification checks will not work correctly if there is any leading or
trailing whitespace on any of the register statement lines or the peer
section headings.

This plugin does not work with Realtime.

This plugin calls the asterisk executable directly, so make sure that the user
executing this script has appropriate permissions!  Usually the asterisk binary
can only be run by the asterisk user or root. To grant the nagios user
permissions to execute the script, try something like the following in your
/etc/sudoers file:
  nagios ALL=(ALL) NOPASSWD: /path/to/plugins/directory/check_asterisk_peers

Then call the plugin using sudo:
  /path/to/sudo check_asterisk_peers
"
}

print_help() {
    print_usage
    echo "Check asterisk peers plugin for Nagios."
    echo ""
}

# Sets the exit status for the plugin.  This is done in such a way that the
# status can only go in one direction: OK -> WARNING -> CRITICAL.
set_exit_status() {
	new_status=$1
	# Nothing needs to be done if the state is already critical, so exclude
	# that case.
	case $exitstatus
	in
		$STATE_WARNING)
			# Only upgrade from warning to critical.
			if [ "$new_status" = "$STATE_CRITICAL" ]; then
				exitstatus=$new_status;
			fi
		;;
		$STATE_OK)
			# Always update state if current state is OK.
			exitstatus=$new_status;
		;;
	esac
}

# Ensures that a call to the Asterisk process returns successfully.  Exits
# critical if not.
check_asterisk_result() {
	if [ "$1" != "0" ]; then
		echo "CRITICAL: $2"
		exit $STATE_CRITICAL
	fi
}

# Recurse through the configuration file and find all included files, building
# a final list of files to use for verification checks.
get_config_files() {
	for file in $1
	do
		if [ "$conf_file_list" ]; then
			conf_file_list="$conf_file_list $file"
		else
			conf_file_list="$file"
		fi
		files=`cat $file | grep ^\#include | awk '{print $2;}'`
		if [ "$files" ]; then
			get_config_files "$files"
		fi
	done
}

# Defaults.
exitstatus=$STATE_OK
peers=
registrations=
verify_peers=
verify_registrations=
peer_type=
conf_file=
conf_file_list=
all_config=
test_ok=
test_errors=
output=

# Grab the command line arguments.
while test -n "$1"; do
    case "$1" in
        --help)
            print_help
            exit $STATE_OK
            ;;
        -h)
            print_help
            exit $STATE_OK
            ;;
        --peers)
            peers=$2
            shift
            ;;
        -p)
            peers=$2
            shift
            ;;
		--registrations)
			registrations=$2
			shift
            ;;
		-r)
			registrations=$2
			shift
            ;;
        --verify-peers)
        	verify_peers=1
        	;;
        --verify-registrations)
        	verify_registrations=1
        	;;
        --type)
            peer_type=$2
            shift
            ;;
        -t)
            peer_type=$2
            shift
            ;;
        --conf-file)
        	conf_file=$2
        	shift
        	;;
        -x)
            exitstatus=$2
            shift
            ;;
        --exitstatus)
            exitstatus=$2
            shift
            ;;
        *)
            echo "Unknown argument: $1"
            print_usage
            exit $STATE_UNKNOWN
            ;;
    esac
    shift
done

# Make sure we have a valid peer type.
if [ "$peer_type" != "" ] && [ "$peer_type" != "sip" ] && [ "$peer_type" != "iax" ]; then
	echo "Peer type must be one of: sip, iax."
	exit $STATE_UNKNOWN
# Convert iax type into internal iax2 type for the Asterisk executable.
elif [ "$peer_type" = "iax" ]; then
	peer_type=iax2
# Default type is SIP.
else
	peer_type=sip
fi

if [ ! "$peers" ] && [ ! "$registrations" ]; then
	echo "At least one peer or registration must be provided."
	exit $STATE_UNKNOWN
fi

# Set a default configuration file if none was set.
if [ "$conf_file" = "" ]; then
	if [ "$peer_type" = "iax2" ]; then
		conf_file=/etc/asterisk/iax.conf
	else
		conf_file=/etc/asterisk/sip.conf
	fi
fi

# Check to make sure the registrations are in the username@uri format.
# This is a weak check, but at least enough to make sure we have something
# in the user/uri variables.
if [ "$registrations" ]; then
	for r in $registrations
	do
		REGISTRATION_FORMAT_CORRECT=`echo "$r" | grep ".\+@.\+"`
		if [ ! "$REGISTRATION_FORMAT_CORRECT" ]; then
			echo "Registration $r is not in the valid username@uri format."
			exit $STATE_UNKNOWN
		fi
	done
fi

# Load the configuration files if necessary.
if [ "$verify_peers" ] || [ "$verify_registrations" ]; then
	get_config_files $conf_file
	all_config=`cat $conf_file_list`
fi

# Check peers.
if [ "$peers" ]; then
	for p in $peers
	do
		if [ "$verify_peers" ]; then
			peer_verified=`echo "$all_config" | grep "^\[${p}\]$"`
		else
			peer_verified=1
		fi
		if [ "$peer_verified" ]; then
			# Fetch the data from asterisk.
			command_output=`$ASTERISK -rx "$peer_type show peer $p" 2>&1`
			check_asterisk_result $? "$command_output"
			status=`echo "$command_output" | grep "^[[:space:]]*Status[[:space:]]*:" | awk '{print $3;}'`
			if [ "$status" = "OK" ]; then
				if [ "$test_ok" ]; then
					test_ok="${test_ok}, $p"
				else
					test_ok="$p"
				fi
			else
				if [ "$status" ]; then
					status_error="$status"
				else
					status_error="Not found"
				fi
				if [ "$test_errors" ]; then
					test_errors="${test_errors}, ${p}: $status_error"
				else
					test_errors="${p}: $status_error"
				fi
			fi
		else
			if [ "$test_ok" ]; then
				test_ok="${test_ok}, $p (not active)"
			else
				test_ok="$p (not active)"
			fi
		fi
	done
fi

# Check registrations.
if [ "$registrations" ]; then
	# Fetch the data from asterisk.
	command_output=`$ASTERISK -rx "$peer_type show registry" 2>&1`
	check_asterisk_result $? "$command_output"
	for r in $registrations
	do
		user=`echo $r | cut -f 1 -d "@"`
		# Add backslashes to escape the dot.
		uri=`echo $r | cut -f 2 -d "@" | sed -e 's/\./\\\./g'`
		if [ "$verify_registrations" ]; then
			# This regex isn't perfect, but it does the trick ok.
			registration_verified=`echo "$all_config" | grep "^register => ${user}:.\+@${uri}$"`
		else
			registration_verified=1
		fi
		if [ "$registration_verified" ]; then
			# Have to cut off registration name at 12 characters, so let's hope
			# it's still unique!
			cut_user=${user:0:12}
			# This regex isn't perfect, but it does the trick ok.
			status=`echo "$command_output" | grep "^${uri}:[[:digit:]].\+[[:space:]].\+${cut_user}[[:space:]].\+" | awk '{print $4;}'`
			if [ "$status" = "Registered" ]; then
				if [ "$test_ok" ]; then
					test_ok="${test_ok}, $r"
				else
					test_ok="$r"
				fi
			else
				if [ "$status" ]; then
					status_error="$status"
				else
					status_error="Not found"
				fi
				if [ "$test_errors" ]; then
					test_errors="${test_errors}, ${r}: $status_error"
				else
					test_errors="${r}: $status_error"
				fi
			fi
		else
			if [ "$test_ok" ]; then
				test_ok="${test_ok}, $r (not active)"
			else
				test_ok="$r (not active)"
			fi
		fi
	done
fi

if [ "$test_errors" ]; then
	output="$output ERROR: $test_errors"
	set_exit_status $STATE_CRITICAL
fi

if [ "$test_ok" ]; then
	output="$output OK: $test_ok"
fi

if [ ! "$output" ]; then
	echo "UNKNOWN: No output"
	exit $STATE_UNKNOWN
fi

echo "$output"

[Ruža Becelin]

ASTERISK="/usr/bin/sudo /usr/sbin/asterisk"

Zkus, jestli to pomuze...

[Dalibor V.]

To nee.. protoze pak to po me chce sudo heslo.. to bzch musel mit v sudoers primo /usr/sbin/asterisk a to nechci ... takze zmenou obsahu scriptu cesta nevede..

[Dalibor V.]

Jen pro klid duse: tohle taky nepomohlo

[linuxik]

Ahoj,

nemas nahodou zapnuty selinux, nebo neco podobneho?

[Dalibor V.]

Ahoj, predpokladam, ze ne

[linuxik]

No tak to je smula, reseni by bylo jednoduchy. Jeste me neco napadlo, kdyz spustis check z Centreonu, tak se spousti pod uzivatele WWW serveru a ne nagiosu.

[Dalibor V.]

No jo, ale porad mu to sudo nekam musim napsat.. sudo y centreonu pred script nejde jen tak napsat..

[linuxik]

No a v cem je problem, napis ho primo do skriptu, pred prikaz ktery potrebuje roota, napr. ja takhle resim test smb sdileni:

sudo mount -t cifs //$HOST/$SHARE ......

funguje to bez problemu, jen musis v sudoers zakomentovat radku "Defaults requiretty" a pro Nagios nastavit spousteni prikazu bez hesla:

nagios ALL=(ALL) NOPASSWD: ALL

[Dalibor V.]

No slo mi spis o to, aby nagios (resp. apache www-data) nemel plnej pristup k asterisku... to je trosku bezp. riziko, neee?

[komodo]

A takhle by to neslo ?

/usr/bin/sudo $USER1$/check_asterisk_peers -p 100

nezkousel jsem, je to jenom napad

[Dalibor V.]

to je prvni, co jsem zkusil.. a to prave nejde :-/ ale diky

[komodo]

A co to poustet pres nrpe ? Jasne je to zbytecny meziclanek, ale asi to bude nejjednodussi.

[David H.]

Také jsem se s tím trochu trápil a nakonec jsem to vyřešil tak, že jsem si udělal nový soubor, který volám z Centreonu a obsahuje to sudo.

Obsah souboru vypadá nějak takto:

#! /bin/sh

/usr/bin/sudo /usr/local/nagios/libexec/check_mailq -M 'qmail' -w '80' -c '100'

[Dalibor V.]

Tak jsem si ve virtualu nainstaloval cistou instalaci Debianu s nagiosem a centreonem (stejne verze). Grafy i pole ARGs funguji bez problemu. Postupoval jsem podle stejneho navodu. Dala by se tahle instalace nejak vyuzit pro rozchozeni meho serveru?

[Dalibor V.]

>> nevykresluji se grafy (resp. necitelny png obrazek) toto VYRESENO: problem s Centreonem a PHP 5.3 (funkce split je deprecated). Zeditoval jsem /etc/php5/apache2/php.ini a zmenil error_reporting = E_ALL & ~E_NOTICE na error_reporting = E_ALL & ~E_NOTICE & ~E_DEPRECATED podle navodu.

[Dalibor V.]

VYRESENO timto i toto: >> v konfiguraci služeb mi chybí pole ARGS pro pridani promennych parametru k prikazu

[Dalibor V.]

nicmene je to divny, protoze v tom virtualu jsem mel stejnou verzi Centreonu i PHP :-/