@bhtooefr This has been solved quite efficiently in Distributed Hash Table protocols https://en.wikipedia.org/wiki/Distributed_hash_table where you can quite efficiently search a large number of nodes for content without a single directory index.