Open adzialocha opened 10 months ago
Out of the top of my head (no laptop with me right now) some rough ideas, probably need improvement:
Connections
/ Peers
etc. in the global Context
, similar to schema managerSchemaIdSet
(collection queries and node announcements) or DocumentId
(document queries)
Only start new replication session when we have new data for another node.
The logic behind this is also needed for client subscriptions.