On 5/7/09 10:53 PM PDT, "Rob Wultsch" <[email protected]> wrote:
> > Thinking about it more I wonder if forcing the transaction to commit > on the slave in the order that they were committed on the master would > be part of the solution. > In fact this is why replication tends to be single-threaded on the slave. It's the only tractable implementation for the general case. Based on my own experience at least, no matter what tack you take to eliminate deadlocks and/or races they pop up elsewhere. That's a characteristic of problems where there is an underlying impossibility result. There may be one here though I have not seen a proof of it. Cheers, Robert > -- > Rob Wultsch > [email protected] > > _______________________________________________ > Mailing list: https://launchpad.net/~drizzle-discuss > Post to : [email protected] > Unsubscribe : https://launchpad.net/~drizzle-discuss > More help : https://help.launchpad.net/ListHelp > _______________________________________________ Mailing list: https://launchpad.net/~drizzle-discuss Post to : [email protected] Unsubscribe : https://launchpad.net/~drizzle-discuss More help : https://help.launchpad.net/ListHelp
