Hello All,
We have just finished implementing our Solarwinds NPM/SAM/NTA/NCM server. I have been working with a vendor with this build, but we ran into an issue that I'm not sure if there is a solution to.
Instead of casting a wide net and doing a network discovery and resulting cleanup, we have decided to start using Solarwinds by monitoring IT Services - trying to use ITIL methodology.
We have a critical application that we decided to start with. The idea is to be able to monitor the application, and all supporting infrastructure that the application depends on. For example, the application is installed on multiple nodes performing different functions. These nodes happen to be virtual machines running on our vmware production cluster. The vmware hosts run on server blades that are installed in a chassis. The chassis is connected to network and SAN switches.
Like many larger shops, our production vmware cluster utilizes the high availability functionality and can dynamically move a virtual machine from one virtual host to another. Configuring solarwinds to monitor a virtual machine is easy. What I cannot figure out...is how to ONLY monitor the virtual hosts that the virtual machines are actually running on at that moment. I cannot find any intelligence that can do this. We have decided to monitor ALL of our production virtual hosts - using custom groups to nest them into a group. The group is a subgroup of the 'IT Service'. The issue with this is that if one of the vmware hosts triggers an alert (turns yellow/red) then the 'IT Service' will turn red - even if the virtual host that is having an issue is not currently running any of the nodes that the service is running on.
If I can figure this out then I would dynamically map only the chassis that the virtual hosts running the virtual machines running the service were running on...and so on.
If anyone has any ideas - they would be welcome! I am somewhat new at this but have found nothing so far in my search for a solution for this issue.