Automatic monitoring

[jkinning]

We currently have a process in place that our Server Management group uses which keeps tracks of several misc items when provisioning new servers. You basically go to a webpage and key in the new server hostname, OS type (Windows, Linux, AIX, etc), if it is a production server or non-production server, if it needs to be backed up, who owns the server, IP address that is assigned to the server, and finally if it needs Nagios monitoring. Today, the Nagios monitoring is manual and I am wondering if I baked the Nagios agent into our Server image build or better yet use are Server Management tool, Tivoli Endpoint Manger from IBM, to deploy the Nagios client is there a way I can get the basic monitoring going?

For example, I stand up a new Windows server called myserver. All the servers will have CPU, Memory, C and D drives monitored. Is it possible to have these basic items monitored automatically or with scripts on the backend to eliminate the risk of whoever is provisioning this new server forgetting to set it up? So basically, whenever a new server is stood up and Tivoli deploys the agent then Nagios XI will start monitoring these basic settings. Once the application gets installed or any additional monitoring is needed the person can go into the system and add the extra items that need to be monitored, services, processes, etc.

I am still relatively new to Nagios XI and have been the sole administrator of the Nagios core system we are running today. I get all these requests and have to manually configure everything so with Nagios XI I am looking for as much automation and shared responsiblity as I can get. How are you using Nagios XI? He is what I am using today but this will change in a few months after I figure out the proper, and correct way to implement Nagios XI.
Nagios 3.0.6
Copyright (c) 1999-2008 Ethan Galstad (http://www.nagios.org)
Last Modified: 12-01-2008
License: GPL

Reading configuration data...

Running pre-flight check on configuration data...

Checking services...
Checked 4334 services.
Checking hosts...
Checked 1126 hosts.
Checking host groups...
Checked 59 host groups.
Checking service groups...
Checked 4 service groups.
Checking contacts...
Checked 513 contacts.
Checking contact groups...
Checked 309 contact groups.
Checking service escalations...
Checked 51 service escalations.
Checking service dependencies...
Checked 0 service dependencies.
Checking host escalations...
Checked 167 host escalations.
Checking host dependencies...
Checked 2 host dependencies.
Checking commands...
Checked 90 commands.
Checking time periods...
Checked 39 time periods.
Checking for circular paths between hosts...
Checking for circular host and service dependencies...
Checking global event handlers...
Checking obsessive compulsive processor commands...
Checking misc settings...

Total Warnings: 0
Total Errors: 0

[sreinhardt]

You absolutely can do this in a variety of ways. I would suggest starting out by watching the Activision talk about dynamic EC2 instance monitoring. While I realize that this is probably more than what you are looking for but they gave a great outline of dynamic loading and unloading of hosts and objects within nagios.

http://www.slideshare.net/nagiosinc/nag ... ten-nwctg1
http://vimeo.com/52112155

In addition to that, do you have a set way of monitoring as of now? ie wmi, nsclient, nsca?
Do you have an idea of how detailed you want to get with the automated additions vs manually specifying things beyond the base os and application monitoring?

[jkinning]

Today we are using wmi for Windows and SSH for *nix. Going forward with Nagios XI we plan to utilize the agents which we will deploy with IBM Endpoint Manager.

Base monitoring for Windows will be:
CPU, Memory, Hard disk (C: D:) and then on Production servers and test the application, process, service, etc that is running on the server.

Linux will be similiar:
CPU, Memory, Hard disk, and processes and services that the application is using along with SSH.

So the slide desk stated that the Activision solution was rolled into the product in previous release. If I setup a generic template for the CPU, Memory, hard disks, etc how does that all work? I believe at this time that is the automated monitoring we are looking for is the base OS and applications.

[abrist]

If I setup a generic template for the CPU, Memory, hard disks, etc how does that all work?

Checks can be inherited. In the most basic example, you would add a hostgroup to your "generic" service checks ("generic" as in they are not applied to a specific host, just a hostgroup and are configured in such a way that any member of the hostgroup will work with the check configuration). You then adda host to the hostgroup, inheriting the hostgroup checks. The best part about this setup is that setting up checks for new hosts can be as simple as creating a host with a template and then applying the hostgroup (if the template did not include it already). The biggest downside is that as the check is "generic", you will not be able to adjust check settings (like interval or thresholds) on an individual, check by check basis. Any changes made to the generic check will affect all checks inherited from it.

[Box293]

As abrist said, using hostgroups for your generic service checks is the quickest way to get a host monitored by Nagios.

When using Tivoli Endpoint Manger, all you need to do is setup some automation that will add the host to Nagios and also add it to the hostgroup you use for your generic services. The following document shows how this can be done:
http://assets.nagios.com/downloads/nagi ... gement.pdf

NOTE: There is a lot of helpful "how to" guides in the help In Nagios XI.
Click the Help tab
Under Documentation Guides click Administrator Guide