Bug report

When snapshotting a mysql table with text columns that have a custom converter registered to it, it unable to convert the string to bytes correctly. `AbstractMysqlFieldReader:74` it's using the mysql charset directly e.g. `utf8mb4` instead of converting this to a Java supported Charset. This will always cause a UnsupportedEncodingException to be thrown and caught.

This logic is surrounded in a try catch. which won't cause any operational problems although it does come with a huge performance penalty. In my local testing I noticed a 10x increase in the amount of rows processed in the same amount of time when the proper charset is used to encode it.