This message was deleted.

For anyone else in this situation - [this](https://github.com/harvester/harvester/pull/1913) PR looks like it would solve this particular problem by moving the service to the guest cluster rather than within the harvester cluster, though I might be misunderstanding something here.