The sequence numbers in two databases don’t have any relation to each other, though; if you sync databases A and B, their lastSequenceNumber values won’t necessarily be the same. Due to implementation details they may be fairly close if you use SQLite
storage, but if you use ForestDB storage they’re usually going to be very different. Other factors like filtered replications will throw them off even more.
The answer to the question in the subject is to let the replication run till it goes idle or stops, with no errors.
I wish there were a better answer, but I’ve found that accurate progress measurement comes at the expense of performance and memory footprint. As we’ve optimized the replicator we’ve lost accuracy in tracking progress.
—Jens