NagiosGraph stops graphing at night


#1

I’ve installed NagiosGraph. Everything seemed to work fine. Graphs were updated correctly, representing correct values. So I went home, but when I returned the next morning I only saw those values that were processed before 17:00 the previous day (the time I left work) en 08:00 that morning (the time that I arrived).

When I check nagiosgraph.log it appears that it has been processing performance data all night, but when I view the graphs I see a blank period between 17:00 and 08:00.

My NagiosGraph configuration is as follows:
#nagiosgraph.conf
debug = 5
logfile = /usr/local/nagios/nagiosgraph/nagiosgraph.log
rrddir = /usr/local/nagios/nagiosgraph/rrd
mapfile = /usr/local/nagios/nagiosgraph/map
colorscheme = 1
heartbeat = 600

Does anyone know this problem and its sollution???

With regards,
Maarten
Edited ]


#2

I don’t use it myself, but if debug = 5 turns on debug, maybe you are out of discspace.


#3

This sounds like a configured timeperiod is responsible. You might want to ‘grep “08:00”’ in your nagios config directory.

-mike


#4

First of all, thank for your replies.

Jakkedup:
Free disk space isn’t the problem. I’ve installed Nagios on a brand new server which I formatted. It has about 50 gb of free disk space left. Besides, if the disk would be full, it wouldn’t start logging again after 08:00 the next morning.

mh720:
I’ve checked for every occurence of a time zone in the nagios config files, but nothing seemed to be wrong there.

Any other suggestions? I’m really stuck here.

With the kindest of regards,
Maarten


#5

I’ve selected one host which has this problem and copied all configuration information of that host into this post. The only thing I’ve changed is the E-mail adres, for the obvious reasons.

I don’t think that the problem lies here, because all hosts are created using the same templates and the same data. We just copy-pasted them into the files. But perhaps someone sees an error…?


hosts.cfg:
define host {
host_name host_pc1117
alias pc1117
address pc1117
use host_tpl_geen_notificatie
contact_groups cont_grp_geen_notificaties
}

host_templates.cfg:
define host {
name host_tpl_geen_notificatie
check_interval 1
process_perf_data 1
retain_status_information 0
flap_detection_enabled 0
retain_nonstatus_information 0
active_checks_enabled 1
passive_checks_enabled 0
obsess_over_host 0
check_freshness 0
check_command cmdc_ping_desktops
max_check_attempts 2
event_handler_enabled 0
notifications_enabled 0
notification_interval 0
notification_period time_nooit
notification_options f
contact_groups cont_grp_geen_notificaties
register 0
}

checkcommands.cfg:
define command {
command_name cmdc_ping_desktops
command_line $USER1$/check_ping_desktops -H $HOSTADDRESS$ -w 100.0,20% -c 500.0,60% -p 1
}

time_periods.cfg:
define timeperiod {
timeperiod_name time_nooit
alias Deze tijdsperiode is blanco
}

contactgroups.cfg:
define contactgroup {
contactgroup_name cont_grp_geen_notificaties
alias De contacts in deze contact-group krijgen nooit notificaties
members cont_geen_notificaties
}

contacts.cfg:
define contact {
contact_name cont_geen_notificaties
use cont_tpl_geen_notificaties
alias Deze contact krijgt nooit notificaties
email [email protected]
}

contact_templates.cfg:
define contact {
name cont_tpl_geen_notificaties
host_notification_period time_nooit
service_notification_period time_nooit
host_notification_options n
service_notification_options n
host_notification_commands cmdn_geen_notificatie
service_notification_commands cmdn_geen_notificatie
register 0
}

misccommands.cfg:
define command {
command_name cmdn_geen_notificatie
command_line /bin/echo “Geen notificatie verzonden” >> /tmp/geen_notificaties.txt
}


#6

I’ve narrowed the problem down to it’s cause (I think at least). The problem is with RRDTool. According to its log the NagiosGraph plugin gives correct input throughout the night. No errors were logged for writing to the database. Only when we read the data out of the RRD database, a lot of entries say ‘nan’ (not a number). So the problem lies with RRDTool I think.

Any suggestions for this?


#7

Are you using the default ‘map’ file for nagiosgraph? This could be that the default map file entry for pings doesn’t like your system’s ping output much.

-mike


#8

sounds to me, that you may want to look at the output from nagios at night, as it is not giving you the same output during the day. Your regex sounds like it not grabbing the data you want therefor, and so you get nan. It could just be, that the device is powered off at night, and so it is not able to ping, so you get nan. Is the device off at night?