A couple good tests are.
1. Corrupt a relay log. For this you can stop the sql thread, cat
/dev/urandom over the newest relay log, start the sql thread and watch
2. Change the innodb_log_file_size in my.cnf without going through the
proper procedure to remove the old log files. In 5.0 this will cause
incorrect information in frm file errors for queries which will take a
little bit of work to track down.
3. Can some random data over myisam files and run a check tables so
mysql marks them as crashed.
On Sun, Apr 18, 2010 at 11:13 AM, Rob Wultsch <wultsch@stripped> wrote:
> On Sun, Apr 18, 2010 at 11:07 AM, Jim Lyons <jlyons4435@stripped> wrote:
>> You can remove the innodb logs and/or the innodb data file. You can also
>> remove some of the individual .idb files (if you're using file-per-table
> He originally asked about how to provide a training excise about
> repairing a db. How the hell do you repair from not having data files?
> For that matter the recovery from lacking log files (and assuming a
> crashed state) is imho ugly as hell.
> Rob Wultsch
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe: http://lists.mysql.com/mysql?unsub=1