On Tuesday 09 November 2004 11:49 am, ian douglas wrote: > My bad, I wasn't running "START SLAVE" to get them going. Seems odd that > this has to be done manually.
I never do.. once its started, it should be fine.. Your just getting errors making it stop. > Also, overnight, my slaves reported this error: > > 041108 19:06:42 [Note] Slave I/O thread: connected to master > '[EMAIL PROTECTED]:3306', replication started in log > 'masterdb-bin.000003' at position 7803960 > 041108 19:06:42 [ERROR] While trying to obtain the list of slaves from > the master 'masterdb.org:3306', user 'repl_user' got the following > error: 'Access denied; you need the REPLICATION SLAVE privilege for this > operation' > 041108 19:06:42 [ERROR] Slave I/O thread exiting, read up to log > 'masterdb-bin.000003', position 7803960 This is what i have: GRANT REPLICATION SLAVE ON *.* TO 'repl'@'%.smelser.org' IDENTIFIED BY PASSWORD 'taken out'; > If a slave dies, is there any way to determine that state from an > external point of view (ie: via Perl/DBD) so we can write a script of > some kind to either connect to the slave and issue a "START SLAVE" or > send the last xx lines of the .err log file to our tech staff? I do this for the err files. It only send them if there is any data.. I use logrotate.. /var/log/mysql/mysql.err { missingok mail [EMAIL PROTECTED] mailfirst daily notifempty compress rotate 2 } /var/log/mysql/mysqld.err { missingok mail [EMAIL PROTECTED] mailfirst daily notifempty compress rotate 2 } Jeff
pgp7Axkk8RKjY.pgp
Description: PGP signature