When ignoring a node/interface/volume in network discovery you need to ignore the node for each polling engine. An option to ignore the node/interface/volume for all current/future polling engines would ease administration.
are you saying you have a machine being monitored by multiple polling engines? That's possible, but a-typical for all but a few special cases (such as if you have pollers in different locations, and want to triangulate up/down of a single system from each location).
Otherwise, nodes (and their associated disks and interfaces) are assigned to a single polling engine.
If you try to add it again to the SAME poller, it's simply skipped
If you try to add it to a different poller, you get a notification during with a prompt about whether you really want to add it or not.
In our environment we move nodes to different pollers for maintenance and troubleshooting activities and to rebalance our polling engines. The initial discovery will be done on the primary polling engine and then moved to one of out additional 7 pollers. On the new polling engine we need to ignore the volumes/interfaces again. It would help us manage our thousands of nodes by being able to flag the volume/interface/node for all pollers at the time of the initial discovery.
