Symmetrically to the archiver module, the indexer module needs to update its data references (content sha1 for now) when new contents are added.

Leveraging the journal infrastructure to have a journal client that inserts new references for indexation in the index queue (orchestrator_all) sounds the way forward.

Related T494