Comment 3 for bug 1100557

@Raghavendra D Prabhu: Ah, that may actually provide a solution to what I'm seeing, I'll try that out and report back. Whether or not it works, I feel that the tool should perhaps be a little more proactive in this scenario and set the variable automatically since some of the operations the tool does can cause some nasty locking in a highly concurrent environment.

@Daniel Nichter: There are 2 "lock wait timeout" variables in play in the server. I experienced this problem on Percona Server 5.5.13-20 and found the following DEFINE in source is what "lock_wait_timeout" is set to: http://bazaar.launchpad.net/~percona-core/percona-server/5.5/view/head:/Percona-Server/sql/sql_const.h#L225

This variable doesn't seem to appear anywhere in the documentation. The other variable is "innodb_lock_wait_timeout" which is set to 50s by default (as per the documentation). I guess I'm curious if the tool is setting "innodb_lock_wait_timeout" and then watching "lock_wait_timeout" to determine when it should timeout.