Details
-
Improvement
-
Status: Resolved
-
Minor
-
Resolution: Abandoned
-
None
-
None
-
None
Description
Presently, we wait for a WAL to be closed before we start replicating the data contained within.
In an "active" system (one that is continually ingesting a steady stream of data), you don't really notice this lag. In a "fresh" system, there is a noticeable wait before you start to get any replication flowing. This tends to be a pain when you want to verify the correct configuration before walking away for a coffee.
I don't believe there is anything that would break from removing this check, but it will require a little testing to make sure the edge cases are solid.