Friday 20 July 2012 11:21:21 pm - 3 replies
I'm running in to a situation with a slave database configured, and wondering if anyone else has seen it before.
I have a master/slave database configuration, with MySQL replication from the master to the slave.
Some times, though, when I publish content objects, I get a failure from eZ. The failure occurs after the object has been instantiated, when the code tries to retrieve the initial version.
I'm guessing this means that the object was INSERT'ed on the master, but the read for the version is SELECT'ed from the slave before the replication has completed from the master.
Has anyone else seen this? If so, is there a good way around it?
Saturday 21 July 2012 10:48:50 pm
it used to be a problem in earlier versions of ezp. I haven't seen the ezp core causing errors in a master/slave setup for a long time.
It's more likely that custom code like a publishing workflow event is causing the issue. In that case you want to wrap some API function calls in a DB transaction.
Another possible reason is TripleM - but it doesn't sound like you're using it.
You must be logged in to post messages in this topic!