Скачать или смотреть Crash-safe semisync replication - Andrei Elkin - MariaDB Server Fest 2021

This video covers recent improvements made to semisync replication recovery.

When a crashed former semisync master recovers from its binary log, the latter may contain events for transactions that were prepared by the storage engine and written to the binary log, but that were never actually received by a replica, or committed by the storage engine.

The crashed master was used to commit those transactions and thus became inconsistent with its former replicas should those servers meanwhile have switched over to a new primary.

The solution presented addresses that through refining the server recovery in such a way that when the old master is restarted in semisync slave mode, past recovery, its binlog will not have any extra transactions.

By Andrei Elkin, who leads the MariaDB replication team at MariaDB Corporation