MySQL: Need to figure out how to handle binlog retention for large table addition initial loads

[Amogh-Bharadwaj]

A MySQL initial load during table addition could take an amount of time greater than the configured value of binlog retention hours on the source MySQL.

In this case, when CDC resumes after the table addition completes, the binlogs we need no longer exist and thus CDC fails, and a resync is the only option.

We need to figure out a way to tackle this phenomenon.

[serprex]

Fix would also help postgres avoid slot lag: start reading sync batches right away