Ever since we removed the crash-safe replication parameters from the slave instance, the sporadic duplicate key issue has never happened again. These three were the only replication parameters that I changed while upgrading to 5.6 so the thought was that they likely caused the issue.
Just wanted to update this ticket.
Ever since we removed the crash-safe replication parameters from the slave instance, the sporadic duplicate key issue has never happened again. These three were the only replication parameters that I changed while upgrading to 5.6 so the thought was that they likely caused the issue.
CHANGED FROM:
relay-log-recovery = 1 info-repository = TABLE info_repository = TABLE
relay-log-
master_
TO:
relay-log-recovery = 0 info-repository = FILE info_repository = FILE
relay-log-
master_