Service configuration best practices

[MiguelAlves]

Hello,

I am a new user of Nagios XI and I am starting my hosts configuration through the Web interface. Initially I started by configuring each host by using the wizards, but after reading your Best practices document, I understood that much probably the best way and more efficient way of configuring my hosts would be by creating host groups and assigning the hosts to these host groups.

Regarding the services configuration, also according to the nest practices document I understood that the best thing would be to configure a service and add it to the host group, and by doing that I would be assigning the service to all the hosts belonging to this host group.

As I understood, this would make the things easier, if I wanted to change the configuration of a specific service, I would only do it in one place and the change would apply to all hosts in that host group, avoiding the need to go into each host and update the service configuration.

Assuming that I will do not have the Bulk Modification Tool, since I will be using the Standard Edition and case I want to modify a service configuration only for one or two particular hosts(override the host group configuration), is there anyway I can achieve it or the only way of having exceptions on service configuration is by adding the services directly to the hosts?


Thanks in advance.
Nuno

[jbrunkow]

Hello @MiguelAlves!

Welcome to the forum.

Yes, your configuration will probably be much neater and more organized if you implement host groups right at the outset. I think it's great that you're starting out by reading the best practices!

Unfortunately, you can not change the configuration for just one service in a host group without adding a new service.

I hope that answers your questions! Let me know if I can provide any other information.

[MiguelAlves]

Thanks for your feedback.

If it is not possible to override the configuration of a service, for a specific host, belonging to the host group, this is probably not the best solution for us.

We identified already one host, that belongs to a Linux Servers Host Group and that needs a special/different, warning and critical alert treshold definition for the memory check service that was assigned to that group.

Isn't there a different way of configuring a service that, would allow us to set a general treshold for the warning and critical alerts, that would apply to the majority of the hosts and then define a specific warning and critical alert treshold value for a specific host? By using Templates or by other means?

Thanks

[scottwilkerson]

The only way to do this would be to exclude this host from the service that contains the hostgroup, see:
https://assets.nagios.com/downloads/nag ... groups.pdf

Then create a new service description for just this host with the different thresholds

[MiguelAlves]

Sorry for the late reply on this topic.

Thanks for your hint. I believe this was what I was looking for.

I have another question:
As stated before I am trying to configure the services individually and then assigning the service to a hostgroup, so that all the machines of that hostgroup can have the same service configuration and so that in the future, case we need to change a service configuration we would only do it in one place.

My question is, in you opinion what would be the most efficient way of configuring a service for checking the inbound bandwidth of some Windows and Linux machine Ethernet interfaces, where we may have more than one ethernet interface and where the name of the interface, may be different from machine to machine. Is there a way of configuring a Service Template leaving the interface name as a variable/argument an then set this argument at the host level configuration or something like this?

Thanks in advance

[scottwilkerson]

My question is, in you opinion what would be the most efficient way of configuring a service for checking the inbound bandwidth of some Windows and Linux machine Ethernet interfaces, where we may have more than one ethernet interface and where the name of the interface, may be different from machine to machine. Is there a way of configuring a Service Template leaving the interface name as a variable/argument an then set this argument at the host level configuration or something like this?

Really the most efficient way to do this is to have separate configurations for each service. In the future if you need to change thresholds in bulk or something, if you have the enterprise version of Nagios XI you can use the Bulk Modification tool in the Core Config Manager

[MiguelAlves]

Thanks for your reply!

Actually I was trying to avoid having to configure individual services for each host, since we are going to use the Standard Edition (no Bulk Modification Tool).

I tried to configure a Service Template, and I tried to use the NAGIOS macros or service/host variables to set the Ethernet Interface name in the Service template and then setting the correct value at the Host level:
check_xi_ncpa! -t $USER6$ -P 5693 -M 'interface/$ARG2$/bytes_recv' -d -u M -w '10' -c '100'
or
check_xi_ncpa! -t $USER6$ -P 5693 -M 'interface/$_SERVICEETHERNAME$/bytes_recv' -d -u M -w '10' -c '100'

...but I had no success!

Would this be possible?

[scottwilkerson]

check_xi_ncpa! -t $USER6$ -P 5693 -M 'interface/$ARG2$/bytes_recv' -d -u M -w '10' -c '100'
or
check_xi_ncpa! -t $USER6$ -P 5693 -M 'interface/$_SERVICEETHERNAME$/bytes_recv' -d -u M -w '10' -c '100'

In both of these scenarios you would have to have a service configuration for each host as that would be where it would pull the $ARG2$ or $_SERVICEETHERNAME$ from

[MiguelAlves]

Hello,

I have already configured some of the Hosts and one switch, following these advises and everything seems to be Ok, for now... Thanks

I have some further questions, regarding Server Redundancy. Case you think it would be better, I can move this to a new Topic...

As far as I understood the Nagios License includes the installation of one Master Server, one Slave Server and one Test Server. My questions are:
- How can we test, during the Trial, this Redundancy (Master and Slave Server)?
- How do we install this Redundant Server?
- Is there any Failover or Failback or High Availability solution we can configure, out of the box or that is pre installed in Nagios XI?
- If there isn’t, do you have, some sort of Guide or Procedure, we can follow to implement this?

Thanks

[scottwilkerson]

Hello,

I have already configured some of the Hosts and one switch, following these advises and everything seems to be Ok, for now... Thanks

Great!

I have some further questions, regarding Server Redundancy. Case you think it would be better, I can move this to a new Topic...

As far as I understood the Nagios License includes the installation of one Master Server, one Slave Server and one Test Server. My questions are:
- How can we test, during the Trial, this Redundancy (Master and Slave Server)?
- How do we install this Redundant Server?
- Is there any Failover or Failback or High Availability solution we can configure, out of the box or that is pre installed in Nagios XI?
- If there isn’t, do you have, some sort of Guide or Procedure, we can follow to implement this?

Thanks

As this is a divergence from the original topic it would be best to create a new thread..

Locking topic