I’m looking at the logs now. I’m the Android guy and replication is, in large part, handled by the native library, LiteCore. I may need help from them to figure out what is going on. Back to you as quickly as possible.
In the maintime, just FYI:
- It is unlikely that the type of connection affects replication. Not impossible, but not likely. I don’t think LiteCore can even tell what the layer 2/3 connection looks like. It is more likely to be the interruption in the connection that is the issue.
- Resetting the checkpoints causes the replication to re-check the entire dataset, to see what needs to be synchronized. There are two reasons this might work: a) because the checkpoint has, somehow, been corrupted, or b) because it re-establishes the connection, at which point everything proceeds normally.
Back to you sometime today