From: Johan De Meersman Date: December 10 2012 12:33pm Subject: Re: MySQL Replication Error List-Archive: http://lists.mysql.com/mysql/228783 Message-Id: <23575659.571.1355142816597.JavaMail.root@tuxera.be> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable ----- Original Message ----- > From: "N=C3=A9stor" >=20 > I spoke to soon!!! > Here is the error about 1.5 hours after replication has started. > 121205 16:39:51 [ERROR] Slave: Error 'Duplicate entry '3468897' > for Yes, that's what you get for running SQL_SLAVE_SKIP_COUNTER without knowing= what the problem is. Your database is now nicely out of sync, and hopefully it's just the watchd= og. The *proper* way of resolving this is to resync the entire database. You *m= ay* get away with deleting watchdog entry 3468897 and restarting replicatio= n as someone has suggested, but I suspect you'll find another half-million = of them. Look at the Percona Toolkit, specifically mk-table-checksum and mk-table-sy= nc, understand how they work and then apply them. If there's too many diffe= rences by now, your only recourse will be to completely reprovision the sla= ve, though. --=20 Linux Bier Wanderung 2012, now also available in Belgium! August, 12 to 19, Diksmuide, Belgium - http://lbw2012.tuxera.be