TO IDENTIFY ERROS WITH GRAPH, verify if:
service-perfdata file exist in Nagios “var” directory ( /var/local/nagios/
# ll /usr/local/nagios/var/service-perfdata
To find out where is inside centreon: Configuration>Nagios>Centreon>Central>”Perfdata file”
service-perfdata is growing (if Nagios insert data il file) (if not try deleting and creating again the file)
service-perfdata is purging by centstorage (presence of service-perfdata_read file)
# tail -f /usr/local/nagios/var/service-perfdata
censtorage processus is running (/etc/init.d/centstorage status)
# ps -ef |grep centstor
no errors in centstorage.log file in Centreon “log” directory
# tail -f /usr/local/centreon/log/centstorage.log
“centstorage” database is growing (look into mysql)
First, you must activate the “Performance Data Processing Option” in your nagios.cfg and select “process-service-perfdata” command in the “Service Performance Data Processing Command” field.
Secondly, edit the process-service-perfdata script in your nagios’s libexec directory. And note PERFFILE variable value. This is the file where nagios will store the performance datas.
Thirdly, in Centreon configuration ( >> Administration > Options > CentStorage > Options), verify that the “storage folder” exists and have correct rights. Then, lowest in the page, in the “Input treatment options” part, the “perdata” field must have the same path as that indicated by the variable PERFFILE in the previous process “service-perfdata” script.
After, you can verify that the service-perfdata file is created and filled by Nagios. Sometimes deleting this files
and recreating it solves the problem.
Be carreful, Centreon read this file to proccess with the perfdata and remove it every minute (it seems to me)
Only if you have something after the pipe symbol, plugin return perfdata. But the perfdata must have the correct syntax
The process-service-perfdata script should be something like:
# Set perfdata file
PERFFILE=”/usr/local/nagios/var/service-perfdata”
# some parameters passed on command line
TIMET=$1
HOSTNAME=$2
SERVICEDESC=$3
HARDSTATE=$4
SERVICESTATE=$5
PERFDATA=$6
# Write data in perfdata file
/usr/bin/printf “%b” “$TIMETt$HOSTNAMEt$SERVICEDESCt$HARDSTATEt$SERVICESTATEt$PERFDATAn” >> $PERFFILE
Always Remembber:
If you modify nagios & Centreon configuration, export your files to nagios and restart it.
good ref.
http://en.doc.centreon.com/Setup:Graphs
http://en.doc.centreon.com/Troubleshooting:Graphs http://forum.centreon.com/tags.php?tag=graph
Hi!
I want to ask for help,
I have not built graphic + report, although there is Monitoring and hosts and services. It is worth sentreon 2.2 + nagios 3.
service-perfdata is purging by centstorage
# Tail-f / usr / local / nagios / var / service-perfdata – ok
censtorage processus is running
# Ps-ef | grep centstor
nagios 720 1 0 2 o’clock? 00:00:00 / usr / bin / perl-w / usr / local / centreon / bin / centstorage
root 8749 17238 0 10:15 pts / 8 00:00:00 grep centstor
no errors in centstorage.log file in Centreon “log” directory
# Tail-f / usr / local / centreon / log / centstorage.log ok
“Centstorage” database is growing (look into mysql) ok
Dear Ferreira,
i just wanna ask.
As we now, in Cacti , if we enter a address that already consist the SNMP it will show all the host that already registered to this address when we want to create graph.
My question is,
how to implemented like that in Centreon,
I’m sorry if my english is not good to you to understand, but i hope you understand my question,
Thanks..
Regards,
Deddy From Indonesia
And review this post:
http://forum.centreon.com/showthread.php/8461-Centreon-suddenly-stops-graphing-errors-executing-process-service-perfdata?p=78386#post78386
Regards,
Ramses