First are there any best practice docs for using Nagios XI? I've heard two different and conflicting things. I've heard I should use templates whenever possible and avoid the monitoring wizard. Are there any gotchas if I setup all my 1150 hosts using the monitoring wizard? Is there a limit on how many host configurations you can make with the wizard without impacting the monitoring? My Nagios XI instance is running on VMware with 3GB and 1vCPU. Is it better to have more memory or more CPU?
I am trying to make this simple for other members of my team but don't want to hinder performance either. I also was trying to read more about the Core Config Manager to setup checks but not really finding anything, or overlooking the information.
Best Practice
Re: Best Practice
First off for the hardware specs:
http://assets.nagios.com/downloads/nagi ... ements.pdf
I'd get a few more core and some extra RAM in there since it looks like this might be a moderately-sized install. Of course it all depends on the total checks and what type of check they are, but a single core and 3GB seems a bit low no matter how you cut it.
As for the CCM:
http://assets.nagios.com/downloads/nagi ... gement.pdf
The doc covers host management but the knowledge transfers well to services, contacts, etc.
And to the original point of templates, that really is a personal preference. They can certainly speed up configuration if you are making a lot of changes by hand in the CCM, but it also adds a layer of complexity. The wizards do a good job of creating things quickly, but as soon as you need to tweak things the CCM is your friend.
http://assets.nagios.com/downloads/nagi ... ements.pdf
I'd get a few more core and some extra RAM in there since it looks like this might be a moderately-sized install. Of course it all depends on the total checks and what type of check they are, but a single core and 3GB seems a bit low no matter how you cut it.
As for the CCM:
http://assets.nagios.com/downloads/nagi ... gement.pdf
The doc covers host management but the knowledge transfers well to services, contacts, etc.
And to the original point of templates, that really is a personal preference. They can certainly speed up configuration if you are making a lot of changes by hand in the CCM, but it also adds a layer of complexity. The wizards do a good job of creating things quickly, but as soon as you need to tweak things the CCM is your friend.
Former Nagios employee
Re: Best Practice
So, it is all a matter of preference? If I have users who in order to allow them to start adding hosts would use the wizard that would be alright? Even if in the end we wind up having say 5000 hosts being monitored? That would be awesome as that is how I was planning on moving forward. The wizard would allow them to get the hosts in with checks and then if needed I or another trained person could fine tune the checks with CCM.
I will increase my VM to 8GB of RAM and 4vCPU (2 sockets and 2 cores). I just really want a simplified installation while providing other non-technical people the ability to add hosts and services to monitor. Mainly Windows and the reason why I was looking at templates but if bringing in hosts manually one-by-one using the wizard is fine I may go down that path. The hosts currently in Core are incomplete and only monitoring hardware items, CPU, RAM, Disk space so bringing them in one at a time I can add more of the services checks. Before with core we were using WMI and going forward with XI we want to use the NSClient++.
I will increase my VM to 8GB of RAM and 4vCPU (2 sockets and 2 cores). I just really want a simplified installation while providing other non-technical people the ability to add hosts and services to monitor. Mainly Windows and the reason why I was looking at templates but if bringing in hosts manually one-by-one using the wizard is fine I may go down that path. The hosts currently in Core are incomplete and only monitoring hardware items, CPU, RAM, Disk space so bringing them in one at a time I can add more of the services checks. Before with core we were using WMI and going forward with XI we want to use the NSClient++.
Re: Best Practice
As cliché as it sounds, Nagios administration is really both an art and a science. Having a solid config plan is the science, and knowing when to make exceptions is the art.
Former Nagios employee
Re: Best Practice
Right but I am hoping to provide a solution, Nagios XI, to allow our Developers to start creating their own checks for their applications. The easiest method I believe for them is to just use the Wizard. So, my thoughts were to have Server Management create the initial server check via the wizard, can a template be used here? Then the Developer would select the host and reconfigure adding the service check they want or is there a better method for all this that all the Nagios veterans know and can share? 
I am a newbie when it comes to all the workings of Nagios and I simply just want to implement Nagios XI correctly the first time and have a working process to add/remove/modify hosts and service checks going forward.
I also plan to allow the Networking and Communication groups along with Developers the ability to monitor and add notification hosts. Since they would be the ones who really know what needs to be monitored, how it needs to be monitored and who should get notified when it is broke.
I am a newbie when it comes to all the workings of Nagios and I simply just want to implement Nagios XI correctly the first time and have a working process to add/remove/modify hosts and service checks going forward.
I also plan to allow the Networking and Communication groups along with Developers the ability to monitor and add notification hosts. Since they would be the ones who really know what needs to be monitored, how it needs to be monitored and who should get notified when it is broke.
Re: Best Practice
If your developers are going to be developing their own plugins, the pre-defined Wizards won't be much help. They are great for setting up SNMP checks of routers or HTTP checks of websites, but if you have custom plugins involved then you would need to set up the check by hand in the CCM. Templates will be useful for setting all the defaults that don't change often (check interval, time periods, etc) but for things like plugins and arguments you need to set those individually.
If you want to do a simple ping check on the host and add in services later, use the Generic Network Device Monitoring Wizard which will just create the host and a single Ping service attached to it. Then you can add whatever services your developers come up with to that host. "generic-service" is a good template to use for services and will be appropriate for most situations. You can always change the values individually if you need to deviate from what the template provides.
If you want to do a simple ping check on the host and add in services later, use the Generic Network Device Monitoring Wizard which will just create the host and a single Ping service attached to it. Then you can add whatever services your developers come up with to that host. "generic-service" is a good template to use for services and will be appropriate for most situations. You can always change the values individually if you need to deviate from what the template provides.
Former Nagios employee
Re: Best Practice
Sorry for the confusion. Our Developers won't be writing custom plugins, at least not yet. I just want to provide them the ability to login to Nagios XI and add the necessary service checks their application may need along with any database connections they may or may not have. It appears initially that in order to do this they would need Admin access to Nagios XI so I thought if they could reconfigure the host using the wizard than that might work.
Going from a single Nagios admin who manually keyed in all the hosts and service checks, if he knew what to include, to trying to allow the world to add their own hosts and service checks is where I am getting hung-up and not really seeing the benefit of Nagios XI.
Will there be anything in the future to leverage Active Directory for Access roles and notifications?
Guess I just need to read more docs.
Going from a single Nagios admin who manually keyed in all the hosts and service checks, if he knew what to include, to trying to allow the world to add their own hosts and service checks is where I am getting hung-up and not really seeing the benefit of Nagios XI.
Will there be anything in the future to leverage Active Directory for Access roles and notifications?
Guess I just need to read more docs.
Re: Best Practice
Wizards are used to create new hosts/services, not alter existing ones. The developers would need to have CCM access in order to create/alter them. In addition, the reason I brought up plugins is that the applications they are developing sound like they will need some fairly specific checks. If one of our existing plugins can monitor the application, that's great and will save you a few steps. You can pretty much use the existing plugins/wizards to make the check. However if you need to create your own plugin to monitor the app, then our existing wizards can't set that up since they don't know what your plugin does or what parameters it expects. In cases like this you would need to define the command and services manually in the CCM.
Former Nagios employee