Nagiosgraph dont collect data from passive checked hosts

[mkot]

1.

Code: Select all

root@nagios:/usr/local/nagios/libexec# ./check_nt -H <host_ip> -p 12489 -v MEMUSE
Memory usage: total:8025.62 Mb - used: 2135.22 Mb (27%) - free: 5890.40 Mb (73%) | 'Memory usage'=2135.22Mb;0.00;0.00;0.00;8025.62

2.

Code: Select all

root@nagios:/usr/local/nagios/libexec# ./check_nt -H <host_ip> -p 12489 -v CPULOAD -l 5,80,85
CPU Load 1% (5 min average) |   '5 min avg Load'=1%;80;85;0;100

3.

Code: Select all

root@nagios:/usr/local/nagios/libexec# ./check_nt -H <host_ip> -p 12489 -v USEDDISKSPACE -l C
C:\ - total: 39.06 Gb - used: 23.79 Gb (61%) - free 15.27 Gb (39%) | 'C:\ Used Space'=23.79Gb;0.00;0.00;0.00;39.06

[abrist]

Alright. Active checks are using the western decimal notation. Can you pull up a few passive checks from nagios.log and check if their status/perfdata is using commas or decimals?
You may want to change the locale on the nagios server to LANG=en_US.UTF-8.

[mkot]

Passive checks:
Some servers:

Code: Select all

tai@nagios:~$ less /usr/local/nagios/var/nagios.log |grep SQL
[1396562400] CURRENT HOST STATE: SQL;UP;HARD;1;PING OK - Packet loss = 0%, RTA = 0.16 ms
[1396562400] CURRENT SERVICE STATE: SQL;C:\ Drive Space;OK;HARD;1;c:\ - total: 33,36 Gb - used: 25,00 Gb (75%) - free 8,36 Gb (25%)
[1396562400] CURRENT SERVICE STATE: SQL;CPU Load;OK;HARD;1;CPU Load 59% (5 min average)
[1396562400] CURRENT SERVICE STATE: SQL;D:\ Drive Space;OK;HARD;1;d:\ - total: 39,06 Gb - used: 38,11 Gb (98%) - free 0,95 Gb (2%)
[1396562400] CURRENT SERVICE STATE: SQL;E:\ Drive Space;OK;HARD;1;e:\ - total: 27,68 Gb - used: 13,26 Gb (48%) - free 14,42 Gb (52%)
[1396562400] CURRENT SERVICE STATE: SQL;MEM Load;OK;HARD;1;Memory usage: total:8427,76 Mb - used: 3246,35 Mb (39%) - free: 5181,41 Mb (61%)
[1396562400] CURRENT SERVICE STATE: SQL;Serwer PING;OK;HARD;1;PING OK - Packet loss = 0%, RTA = 0.14 ms
[1396562400] CURRENT SERVICE STATE: SQL;Uptime;OK;HARD;1;System Uptime - 112 day(s) 6 hour(s) 25 minute(s)

Code: Select all

tai@nagios:~$ less /usr/local/nagios/var/nagios.log |grep WA-OCR2
[1396562400] CURRENT HOST STATE: WA-OCR2;UP;HARD;1;PING OK - Packet loss = 0%, RTA = 0.25 ms
[1396562400] CURRENT SERVICE STATE: WA-OCR2;C:\ Drive Space;OK;HARD;1;c:\ - total: 97,65 Gb - used: 40,47 Gb (41%) - free 57,18 Gb (59%)
[1396562400] CURRENT SERVICE STATE: WA-OCR2;CPU Load;OK;HARD;1;CPU Load 0% (5 min average)
[1396562400] CURRENT SERVICE STATE: WA-OCR2;MEM Load;OK;HARD;1;Memory usage: total:5201,44 Mb - used: 933,94 Mb (18%) - free: 4267,49 Mb (82%)
[1396562400] CURRENT SERVICE STATE: WA-OCR2;Serwer PING;OK;HARD;1;PING OK - Packet loss = 0%, RTA = 0.35 ms
[1396562400] CURRENT SERVICE STATE: WA-OCR2;Uptime;OK;HARD;1;System Uptime - 21 day(s) 16 hour(s) 6 minute(s)

Some PC/NB hosts:

Code: Select all

tai@nagios:~$ less /usr/local/nagios/var/nagios.log |grep PN-INFO1
[1396562400] CURRENT HOST STATE: PN-INFO1;UP;HARD;1;PING OK - Packet loss = 0%, RTA = 72.68 ms
[1396562400] CURRENT SERVICE STATE: PN-INFO1;CPU Load;OK;HARD;1;OK CPU Load ok.
[1396562400] CURRENT SERVICE STATE: PN-INFO1;DISK Load C;OK;HARD;1;OK: All drives within bounds.
[1396562400] CURRENT SERVICE STATE: PN-INFO1;MEM Load;OK;HARD;1;OK: physical memory: Total: 2.97G - Used: 1.95G (65%) - Free: 1.02G (35%)

Code: Select all

tai@nagios:~$ less /usr/local/nagios/var/nagios.log |grep WA-AMULLER
[1396562400] CURRENT HOST STATE: WA-AMULLER;UP;HARD;1;PING OK - Packet loss = 0%, RTA = 0.13 ms
[1396562400] CURRENT SERVICE STATE: WA-AMULLER;CPU Load;OK;HARD;1;OK CPU Load ok.
[1396562400] CURRENT SERVICE STATE: WA-AMULLER;DISK Load C;OK;HARD;1;OK: All drives within bounds.
[1396562400] CURRENT SERVICE STATE: WA-AMULLER;MEM Load;OK;HARD;1;OK: physical memory: 1.63G

Code: Select all

tai@nagios:~$ less /usr/local/nagios/var/nagios.log |grep WA-ADZIENISZEW1
[1396585619] EXTERNAL COMMAND: PROCESS_SERVICE_CHECK_RESULT;WA-ADZIENISZEW1;MEM Load;0;OK: physical memory: Total: 7.68G - Used: 1.67G (21%) - Free: 6.01G (79%)|'physical memory %'=21%;90;95 'physical memory'=1.65999G;6.91;7.29;0;7.59999
[1396585619] PASSIVE SERVICE CHECK: WA-ADZIENISZEW1;MEM Load;0;OK: physical memory: Total: 7.68G - Used: 1.67G (21%) - Free: 6.01G (79%)
[1396585752] EXTERNAL COMMAND: PROCESS_SERVICE_CHECK_RESULT;WA-ADZIENISZEW1;DISK Load C;0;OK: All drives within bounds.|'C: %'=91%;2;1 'C:'=43.54G;9.27999;4.63999;0;464.13
[1396585752] PASSIVE SERVICE CHECK: WA-ADZIENISZEW1;DISK Load C;0;OK: All drives within bounds.
[1396585926] EXTERNAL COMMAND: PROCESS_SERVICE_CHECK_RESULT;WA-ADZIENISZEW1;CPU Load;0;OK CPU Load ok.|'5m'=6%;50;80 '1m'=6%;50;80 '30s'=5%;50;80
[1396585926] PASSIVE SERVICE CHECK: WA-ADZIENISZEW1;CPU Load;0;OK CPU Load ok.

On Windows XP, 200/2003 Server we're using NSClient++-0.3.9-Win32 and on Windows Vista/7/8 we're using NSCP-0.4.1.90-Win32.

[abrist]

Is there any pattern here? Are the passive checks with commas coming from systems without a western decimal locale?
Are the passive checks without commas coming from boxes configured with a western locale?
Are these boxes using different builds of nsca?

[slansing]

[1396562400] CURRENT SERVICE STATE: PN-INFO1;MEM Load;OK;HARD;1;OK: physical memory: Total: 2.97G - Used: 1.95G (65%) - Free: 1.02G (35%)

[1396562400] CURRENT SERVICE STATE: WA-OCR2;MEM Load;OK;HARD;1;Memory usage: total:5201,44 Mb - used: 933,94 Mb (18%) - free: 4267,49 Mb (82%)

It is curious that one of these is returning a "," the other a "." though they seem to be almost different check formats. You are sure these are both forwarded results from NSCA?

[mkot]

Is there any pattern here? Are the passive checks with commas coming from systems without a western decimal locale?
Are the passive checks without commas coming from boxes configured with a western locale?
Are these boxes using different builds of nsca?

On all hosts locale settings're configuret for Polish settings (we're using "," to split decimal fractions).
Yes, boxes're using different builds of nsca - Win 2000/2003/XP - nsclient 0.3.X, Win Vista/7/8 - nsclient 0.4.X.

[1396562400] CURRENT SERVICE STATE: PN-INFO1;MEM Load;OK;HARD;1;OK: physical memory: Total: 2.97G - Used: 1.95G (65%) - Free: 1.02G (35%)

[1396562400] CURRENT SERVICE STATE: WA-OCR2;MEM Load;OK;HARD;1;Memory usage: total:5201,44 Mb - used: 933,94 Mb (18%) - free: 4267,49 Mb (82%)

It is curious that one of these is returning a "," the other a "." though they seem to be almost different check formats. You are sure these are both forwarded results from NSCA?

All Windows XP/Vista/7/8 hosts're monitorng by NSCA plugin (passive checks), but Windows Server 2000/2003 (+ 1 host with WinXP - WA-OCR2) and Linux Ubuntu, CentOS, IPCop (I also tried configure it on FreeBSD 7.2 but I failed ) are monitoring by NRPE (active checks). I can try enable passive checks on Servers to, maybe it'll help.
On Windows 2000/2003/XP I've installed nsclient 0.3.X and on Windows Vista/7/8 I've installed nsclient 0.4.X. I had to do that because on Windows < Vista nsclient 0.4.X get crashed.


If I change locale at Nagios serwer to en_US.UTF-8 will it help?

[abrist]

If I change locale at Nagios serwer to en_US.UTF-8 will it help?

Looking over your results above, I would have to guess "no". You will need to fix the status results on the remote hosts as your XI server is already receiving both dot and comma results.

[mkot]

So if there is possibility I should change active services checking ( Nagios Core's receiving ",") to passive checking (to get "."). Ok, I'll try to do this...

I've got one more question. Why some of my hosts have "-nan" in Nagios Graph? Don't they send vlues?

[abrist]

I've got one more question. Why some of my hosts have "-nan" in Nagios Graph? Don't they send vlues?

Check the plugin output, this may be related to the dot/comma issue as commas instead of dots in the perfdata could cause parsing errors. If you have many NaNs for services that are returning correctly formatted perfdata strings, it could be due to the checks being infrequent (>30min interval), or the plugin not returning any perfdata (when at one time it did).

[mkot]

Ok, firstly I think it would be good if I disable Nagiosgraph for hosts/services it don't need to be graphs generated. For example PING for dumb hosts (127.0.0.1) is not neccesary . I'll try find howto and disable it for hosts/services I don't want infformations stored for.