On Fri, Jan 10, 2025 at 3:15 PM Simon Avery <Simon.Avery@atass-sports.co.uk> wrote:
Then I restored another copy from the same minute as the split, and after a few hundred HA_ERR_FOUND_DUPP_KEY skips, it started catching up. An hour later, it was 0 seconds behind and has stayed in sync.
Those errors shouldn't have happened. Are you sure you are running the master with synv_binlog=1 and the slave with sync_master_info=1, and both with innodb_flush_log_at_trx_commit=1 ?
sync_binlog is indeed set to 0. The others are fine.
That would indeed explain it. Unfortunately, any setting other than 1 for this means that your binlogs on the master will be inconsistent with the state of the data in the tables in case of a crash (or a snapshot based backup which only gives crash level consistency).