Metrics show only data from SNMP checks

[richelieu]

Hi,
I'd like to use Metrics as Metrics Overview in Dashboards, but it shows very little data.
Most of our services get data from NSClient++ and very little services get data from SNMP or WMI.
It seems that Metrics show data only from SNMP services.
How to use Metrics for data we got from NSClient++?
Is there any special requirements by which Metrics recognize appropriate data (memory, cpu, disc usage, ...)?
Regards,
Marek

[rkennedy]

It should be accepting the standard NSClient++ information - can you show us a sample of the checks that are working in the metrics component, and also a sample of what's not?

[richelieu]

Hi,
I'm sending example of checks that work and that do not work.

Host SGEB-IS003 - check by SNMP - working ok in Metrics Overview:
Output: C:\ Label:System Serial Number 42db35f9: 39%used(23966MB/61440MB) (<80%) : OK
Performance Data: 'C:\_Label:System__Serial_Number_42db35f9'=23966MB;49152;58368;0;61440

Host EU-DC010 - check by NSClient++ - not working in Metrics Overview:
Output: OK c: Total: 58.594GB - Used: 18.057GB (31%) - Free: 40.536GB (69%)
Performance Data: 'c used'=18.05741GB;46.87499;52.73437;0;58.59374 'c used %'=30%;79;89;0;100
Output: OK: CPU Load: 5m: average load 0%, 5m: average load 0%, 30m: average load 0%, 30m: average load 0%
Performance Data: 'core 0 5m'=0%;80;90 'total 5m'=0%;80;90 'core 0 30m'=0%;80;90 'total 30m'=0%;80;90
Output: OK: physical: Total: 4GB - Used: 2.011GB (50%) - Free: 1.989GB (49%)
Performance Data: 'physical'=2.01103GB;3.79958;3.91957;0;3.99956 'physical %'=50%;94;97;0;100
Output: OK: committed: Total: 8GB - Used: 1.783GB (22%) - Free: 6.216GB (77%)
Performance Data: 'committed'=1.78344GB;7.59958;7.83957;0;7.99956 'committed %'=22%;94;97;0;100

Up to now we use old NSClient++ syntax (CheckCPU, CheckMem, ... not the new one: check_cpu, check_mem, ...).

Regards,
Marek

[tgriep]

What version of Nagios XI are you running?
Can you login to the XI GUI and go to the Admin > Manage Components and post the version of the Metrics component that is installed on your server?
Check for any updates on that screen and see if there is an updated Metrics component you can install.

[richelieu]

Hi,
Nagios XI version is 5.4.1.
Metrics component version is 1.2.9.
I cannot run "Check for Updates", as nagios server does not have access to Internet. Is there possibility do download newer version to install it offline?
Regards, Marek

[dwhitfield]

Hi,
Nagios XI version is 5.4.1.
Metrics component version is 1.2.9.
I cannot run "Check for Updates", as nagios server does not have access to Internet. Is there possibility do download newer version to install it offline?
Regards, Marek

There's not an offline installer for 5.4.1. Is that a type of 5.4.0 or did you install 5.4.1 and then pull the plug? Anyway, there is not an update to the Metrics component.

What version of NSClient are you using? Can you post your NSClient.ini for us to view? Are the NSClient checks working otherwise? You mention you are using the old syntax. Did you recently upgrade NSClient? NSClient upgrades generally require cleaning the entire directory rather than installing over the top.

Are the normal graphs for those checks working? I see you have performance data.

You mention WMI. Is that working properly?

[richelieu]

Hello,

as for nagiosxi version - there is offline installer for version 5.4.1 - on page https://www.nagios.com/downloads/nagios-xi/ I click on Older Releases link and there are probably all versions that was released. The only problem I had was, that "tar xzf <file>" was not working and I had to use "gunzip <file>" and then "tar xf <file>". So in the GUI I see "Nagios XI 5.4.1" before the link for checking for updates.

We use client NSCP-0.4.4.23-x64 and some new installation has NSCP-0.5.0.62-x64.
Installation are clean (at least most of them), not upgraded.
By menitonig legacy syntax, I mean that commands passed to check_nrpe are CheckCPU and CheckMem and not new one as check_cpu and check_memory.
I will append the file nsclient.ini, but main changes to defaults are: insecure mode, allow arguments, timeouts and definition how to run some external scripts.
Normal graphs are working ok.
WMI we use only in testing environment, but they seems to work with Metrics ok.

Regards,
Marek

[ssax]

Please PM one of us a copy of your profile, you can download it by going to Admin > System Config > System Profile and click the Download Profile button in the top right corner.

I'd like to compare the commands/services/etc with the backend code to see why they are not being picked up.


Thank you

[avandemore]

XI > Admin > System Profile > Download Profile

Please include the zip file in your response. You can PM myself or other support personnel if you have privacy concerns.

[richelieu]

Hi,
I've sent requested profile as private message.
Regards,
Marek