What do you mean by "resolving two statements with the same ID"? Is this about updating rows with the same PK/UK id?
Does this problem still happen with latest PXC 5.5.37 or 5.6.19? If so, can you also attach full processlist and innodb engine status outputs from the time node or cluster is stalled? Also "show status like 'ws%';" from this and at least other node.
Dimitry, the lock_wait_timeout = 31536000 is just default value since MySQL 5.5+, nothing strange here (http:// dev.mysql. com/doc/ refman/ 5.5/en/ server- system- variables. html#sysvar_ lock_wait_ timeout).
Is this issue maybe related to another bug report of yours: https:/ /bugs.launchpad .net/percona- xtradb- cluster/ +bug/1197771 ?
What do you mean by "resolving two statements with the same ID"? Is this about updating rows with the same PK/UK id?
Does this problem still happen with latest PXC 5.5.37 or 5.6.19? If so, can you also attach full processlist and innodb engine status outputs from the time node or cluster is stalled? Also "show status like 'ws%';" from this and at least other node.