Log sequence in the future means that, for whatever reason, the update in the data pages happened but update in the Innodb's log didn't. The InnoDB by itself, without backups, is not protected against media failures, and this happens to be just that. Innodb_force_recovery is not really a regular recovery method, it is more like a last resort attempt to salvage the database and its ability to save the data is more or less luck.
On Nov 6, 2011, at 2:35 AM, Reindl Harald wrote: > > > Am 06.11.2011 06:05, schrieb Kevin Wang: >> I stopped mysql only to find that it wouldn't come back up, /etc/init.d/mysql >> start only outputs . . . . . . failed. I've narrowed it down to an issue >> with InnoDB. The database starts when innodb_force_recovery = 5 and nothing >> lower. When I "check table" for my MyISAM tables, they check fine, but the >> connection is dropped when I attempt to do so for my InnoDB tables with >> force_recovery set. mysql.err and mysql.log are empty. Here are the >> relevant logs from syslog when I attempt to start with >> innodb_force_recovery set to zero: http://pastebin.com/jzhEuWFu >> >> and here's my my.cnf file: http://pastebin.com/qn6huZ09 > > why do you care about starting mysql as long the systme is whining > the whole time over a dying harddisk? > > What did you think "Unrecovered read error" and "Medium error" > try to tell you? this are kernel messages saying: hopefully you > have a backup and if not now is learning lesson for the future > -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/mysql?unsub=arch...@jab.org
