Comment 8 for bug 1516933

Dom (launchpal) wrote :

Hi Krunal

Ah! It could be made much clearer that deadlock counters/logging is only relevent at the local innodb level. I couldn't use Ramesh's steps to invoke deadlock - I just get Lock wait timeout exceeded, but instead I invoked a local deadlock using innotop (commands D, w). And I do now see the deadlock counter increase and the deadlock logs to error log.

So, is there a way to log the cluster deadlocks? This could be made much clearer in the pxc docs/release notes, I'm not the only one who has made this confusion.