General Question about monitoring Windows Machines

[duostech]

Nagios XI 2011 R2.2, running on CentOS VM Image. I am attempting to monitor several aspects of a mixture of server manufacturers to include Dell, HP and SuperMicro. I have found the Dell plugin, and have experienced moderate success. I would like to get more granular, more visible (rather than a global, 'all good') reporting on aspects such as CPU utilization, hard disk utilization, and hard disk capacity. I have moderate experience with Nagios Core, and was able to successfully monitor drive status and capacity, but am running into configuration issues with XI. I started with http://assets.nagios.com/downloads/nagi ... Client.pdf, http://www.nsclient.org/nscp/wiki/CheckDisk, and a few other guides, but am still having issue with the best methods that some of you have found monitoring systems such as this. Is NSClient++ the best rout to go with for Windows machines, SNMP, WMI, etc?

I do have both SNMP and NSClient++ running on a particular site that I am testing all of this on currently. Can anyone point me to a document that may give greater detail as to how precisely configure these hosts from within Core Config Manager? Is there perhaps a 'best practices' manual that might layout step 1, step 2, step 3, etc? As odd as this is going to sound, I am more accustomed to the text-based config files used in Core, and am trying to find a correlation between the two. Thanks for the patience with a NagiosXI noob

[lmiltchev]

Is NSClient++ the best rout to go with for Windows machines, SNMP, WMI, etc?

Yes, it is. Even though, you can use either of them, I believe that NSClient++ will let you perform the widest variety of checks.

Can anyone point me to a document that may give greater detail as to how precisely configure these hosts from within Core Config Manager? Is there perhaps a 'best practices' manual that might layout step 1, step 2, step 3, etc?

1. You will have to install the agent on the client machine:
http://assets.nagios.com/downloads/nagi ... _Agent.pdf

You may also check these options if relevant for your environment:
http://assets.nagios.com/downloads/nagi ... Client.pdf
http://assets.nagios.com/downloads/nagi ... Checks.pdf

2. You run the Windows Server Monitoring Wizard (Configure->Run the Monitoring Wizard->Windows Server) and set it up as you go. We don't have a detailed document on running the wizard, but it is pretty much straight forward process. You can "tweak" some of the settings later through the Core Config Manager (if you have to), but the initial setup is done through running the wizard.

Hope this helps.

[duostech]

Thanks much. I had read the Windows Agent piece already, but went back through it again, just to be safe. I think the passive checks helped push whatever it was that was holding me back over the edge. Here's what I wound up doing to monitor disks:


That much worked quite well, in addition to the Dell Openmanage Monitoring. I am moving onto trying to monitor CPU & Memory utilization and would like to monitor uptime as well. Going off of: http://www.nsclient.org/nscp/wiki/CheckUpTime, I created a custom command since the base Check_nrpe didn't appear to do what I was looking for.


I applied this custom command to monitored service:



From within the Service Status Window, I see:



and from the commandline, I see:



I know I'm missing a super-simple, basic step here, but after staring at it for hours, and spamming the techs with "I appear to be doing fine" lol... I would like to ask for another set of eyes, please. Hopefully, I can take what you folks suggest and apply it to the Uptime and memory monitoring component as well.

I do have proper ports open on the firewall (ASA), as well as on the local firewall (I've even disabled it to test). The attached config does work for drive monitoring, but not for the other. I do recognize that the primary difference is check_nt vs check_nrpe, but referencing info from the nsclient++ wiki, that appears to be the suggested path.

Thoughts?

[scottwilkerson]

In the check command wrap the args you are sending after the -a in quotes. I can't see the whole thing, but something like

Code: Select all

$USER1$/check_nrpe -H $HOSTADDRESS$ -p 5666 -c CheckCPU -a "warn=100 crit=100 time=1"

If you don't it thinks they are arguments to the check_nrpe, not arguments to be sent to the client

[duostech]

Thanks for the reply. I found that my original command did not have a "-c" in front of the CheckCPU command, and as such check_nrpe thought I was issuing the generic check_nrpe command. Even though I had gone over and over the NSC.ini file, I found that elusive ";" right in front of the "allow_arguments" under the [NRPE] section. I had it properly set in the [NSClient] section, but missed it on the NRPE section. Fixed those issues, restarted services, and all worked well. Now I need to get it to actually report utilization, and not just that it's operating within parameters.

Thanks for all of the help!!!