Nagios Support Forum

Hi,

I would like to create my own auto discovery wizard.
I would like to find systems using nmap then perform a snmp poll of the system object ID and hostname then add the new discovered systems to pre-defined hostgroups.
I do not need to perform a scan of hosts for all available ports, so I am hoping this would make the auto discovery quicker.

Is there any information on how to input new hosts in to the MySQL database properly?

Thanks for any help!

Just to get this out there, this is not something that would be covered by our support SLA so we might end up moving this to the General forums at some point.

That being said, I would think it would be much easier and safer to modify the wizard instead of writing one from scratch. Is there any reason you can't just specify a range like 192.168.1.1-192.168.1.100?

tmcdonald,

It's not about specifying a range. we plan on rolling out a Nagios setup at each of our sites. I want to make it easier on the implementation people and have the auto discovery find systems, then determine what those systems are ( including our own systems with their system object id's) then have pre-built hostgroups with checks already configured for each system type.

On our first install it took several hours to discover a each network and have the systems added to Nagios. I am hoping to slim down the discovery phase to just find systems that are up then try performing and snmp poll of their sysName .1.3.6.1.2.1.1.5.0 and sysObjectID .1.3.6.1.2.1.1.2.0
The current Auto Discovery wizard tries to determine all available ports and services a system has available to monitor. We don't need to do this.

We have several different types of systems with different checks.
SAN systems
Kiosks
Workstations
Servers

I am hoping to slim down the discovery phase to just find systems that are up then try performing and snmp poll of their sysName .1.3.6.1.2.1.1.5.0 and sysObjectID .1.3.6.1.2.1.1.2.0

This is an interesting idea. It's a good candidate for a feature request. We could come up with a totally different wizard or extend the capabilities of the existing "auto-discovery" wizard. There could be some challenges though. Are you going to use the same community string on ALL of your systems?

lmiltchev,

I would have to say all of our systems for the most part use the same community strings which are different for each site. I created checks that use one of the macro definitions, so the site admin would only have to add it to one file.

We have been using Openview for years. One of the things I liked about it was the filter options. I can create filters on systems depending on their hostnames, you can use variables, and options like if it starts with KSK, by IP address ranges, or use the System Object ID to filter them. This lets us filter the systems in the inventory list and in the snmp trap receiver window. In the Openview Auto discovery you can also limit the discovery by system Object ID so only those sysObjID's will be found and added, when your paying per node license you will want to limit how many systems are automatically added.

All of our sites have 90% - 95% of the same setup of equipment. I will have pre-defined checks in a hostgroup template so if auto discovery can determine that its a kiosk the system would be added automatically to the KIOSK hostgroup. Then all of the checks I have written for testing the internal devices will be applied.

I am also trying to figure out the best way of dealing with DHCP hosts. In our large sites that use DHCP systems are removed and re-added hostnames typically stay the same but the IP address can change, I am worried that this could be a issue with Nagios. XI lets you run the auto discovery periodically, but not sure if it can detect that the IP address changed for a hostname already found.

This functionality is not available in XI "out of the box" but I believe it is doable. I would recommend contacting our sales team at [email protected] and requesting a quote for a custom development.