Some FAQs on Discovery -
Q. Network devices that no longer exist on the network - servers that are shut down or deleted, network devices that are decommissioned etc, are they automatically removed from the Nimsoft database after some time or do they hang around forever? With database I mean if they are visible in the NiS Manager in Infrastructure Manager, or in UMP.
A. There is an option in the discovery_server to expire server after a select period of time if they don’t appears on a discovery run.
Q. What happens if a device is decommissioned and the IP address is later reused for another server? Will the previous object/device be deleted or overwritten by the new device? Will both devices exist in the database?
A. If a device is decomm'd and the IP is re-used, it will be re-discovered but with different information (hostname, os type maybe, also the credentials used to log in may/may not be different).
Q. How can I delete devices that have been previously discovered from the database? Ie if the network scope of the discovery_agent is changed, and some parts of the subnet has now been excluded, how do I delete those devices that were previously discovered? It does not seem like the now excluded devices are removed automatically? And it is not possible to delete them manually either?
A. Best way if you know what you're looking for is to find the IP and name of that host and issue a:
delete from cm_computer_system where ip = '<ip of the host>' and name = '<name of the host>'
Obviously, make sure you turn that into a SELECT first so you know what you'll be deleting. A discovery reset is a bit like hitting a fly with a cruise missile. Effective, but it hits way more than you really want it to, and the cleanup can take a while.
Q. How do I handle discovery of devices that are discovered multiple times by the same or different discovery agents? Ie routers, load balancers, devices with VIP addresses, other devices with multiple legs or routers running HSRP. How do I tell Nimsoft that these devices are the same device?
A. Simple answer is don't overlap your scopes. This is probably what the customer is doing if he's discovering the same things multiple times (other than routers, which I admit can be problematic). In the case of things with multiple IP's, discovery_agents will discover each device and the da_id will be different, causing duplication. For now, limiting discovery of those devices to one scope is the best way. The other IP's will be discovered too as long as SNMP credentials are proper for those routers.