Note: The plugin uses only persistence in memory. There is no impact on the Jenkins infrastructure (no new files created).
However, we've seen in practice that all depended-upon artifacts are downloaded to a ivy-trigger-cache directory outside the workspace. I'd expect that there would have to be some XML files downloaded and persisted between checks, but downloading all the depended-upon artifacts is really putting a strain on build machines handling multiple, very large, builds.
I'm not sure if this is a requirement of the Ivy library, or if is really is possible to do a resolve without an implicit download. If it is however, that would be ideal for both my poor aching disks and for the network.