Standardise our services

[jamesc23]

Hi,

We current have several hundred servers with services that were created through monitoring wizard. Since then, lots of services like C: or D: ,etc. have been customised to suit each server's criterias.

Here's what I want to achieve:

- I want to create a template for C: drive and apply this template to a host group so that we have C: drive service with standard settings across all servers.
- I want to remove the C: drive service from all servers that were created using monitoring wizard.
- I want to able to modify C: drive individually on a specific server when required, without affecting C: drive service on other servers

Is the above achievable?
Thanks.

[Box293]

- I want to create a template for C: drive and apply this template to a host group so that we have C: drive service with standard settings across all servers.

So you will need to define one service per drive letter and assign the service to a hostgroup. You could have a hostgroup for servers with C: drives and another for servers with D: drives. Easy. This doesn't really need a template. Just duplicate on of the existing services and call it "windows_common_c" or something like that.

- I want to remove the C: drive service from all servers that were created using monitoring wizard.

The service I told you to create, make sure the "Description" is the same as the one that was created using the wizard. Then when you delete all the individual services the existing performance data will not be lost. You can delete the old services by going into CCM > Monitoring > Services and then search for the services. Use the top left checkbox to select them all and then down the bottom under "With Checked" select Delete and then click Go. Wait while it does this and when done Apply Config. Personally I prefer to Deactivate, Apply, Delete, Apply ... this way if a problem occurs during the first Apply I can just activate again.

- I want to able to modify C: drive individually on a specific server when required, without affecting C: drive service on other servers

That's actually the complicated part, there is no quick and easy way to do this. You can:
a) Remove the server from the "windows_common_c" and then create an individual service again.

b) Remove the server from the "windows_common_c". Create another hostgroup called "windows_common_c_specific" and add the server to that. Create another service that uses this new hostgroup. Now comes the complicated bit. On the HOST object go to the Misc Settings tab and add a Free Variable called _C_WARN and _C_CRIT to define specific warning and critical thresholds. Then in that new service you created, instead of defining the the thresholds in the $ARGx$ field, reference it as $_HOSTC_WARN$ and $_HOSTC_CRIT$. What you have done here is stored the warning and critical thresholds in the host object itself and referenced them from the service. So any host that gets added to "windows_common_c_specific" will need Free Variables _C_WARN and _C_CRIT defined in it's host object.

Here's some reading material on free variables:
https://assets.nagios.com/downloads/nag ... tvars.html

Does this make sense? Does this help?

[jamesc23]

Thanks very much for this. I will start playing with this as soon as I've figured out how to remove old, unused host groups.

I currently have a host group which contains no servers in it. But when I go to individual hosts > common settings > host groups, I can see that a host is part of the mentioned host group and hence this host group can't be deleted until all hosts are removed from it.

Is there a faster way to remove all hosts from this host group within having to go to individual hosts? The other problem is that I can't see which hosts are part of this host group until I apply the changes so that I get the error message telling me the name of the host that's associated to this host group.

Thanks.

[jamesc23]

Another one.... I've got a command which reports on uptime with a warning and critical alerts.

I'm looking at setting warning alert to send out notification once a week starting Monday, and critical alert to send out 1 notification once daily. I don't see options where I can do this within the service. Does this mean that I would have to create 2 commands, one for warning and one for critical and set up alerts on each of them accordingly?

[jdalrymple]

Is there a faster way to remove all hosts from this host group within having to go to individual hosts?

No - I believe this may exist in the next version of XI, but in current versions you can only add hosts to a hostgroup with the bulk mod tool, not remove.

The other problem is that I can't see which hosts are part of this host group until I apply the changes so that I get the error message telling me the name of the host that's associated to this host group.

Hostgroup Summary from the Home screen should lead you to this information.

I'm looking at setting warning alert to send out notification once a week starting Monday, and critical alert to send out 1 notification once daily. I don't see options where I can do this within the service. Does this mean that I would have to create 2 commands, one for warning and one for critical and set up alerts on each of them accordingly?

This sounds like a separate topic, but I'll try to answer quickly. If my answer doesn't resolve please open a new topic.
This question does get asked more regularly than a guy might expect. It should be doable however with service escalations. Set up 2 escalations for your service, one that only notifies on warning at your desired interval and one that only notifies on critical at your desired interval. Documentation:
https://assets.nagios.com/downloads/nag ... ations.pdf

[jdalrymple]

As I mentioned - off-topic replies need to be split off. This makes the forum more useful for everyone. In the future when someone Google searches for "problems with SMS and E-mail notifications" they're quite unlikely to click on a topic titled "Standardise our services"

I have split your topic to a new one here:

https://support.nagios.com/forum/viewto ... 441#150527

If the problem in this topic is indeed resolved, can I lock it?

Thanks.

[jamesc23]

Please close it. Thanks.