On Sun, Sep 14, 2008 at 10:27 AM, Jeremy Zawodny <[EMAIL PROTECTED]> wrote: > On 9/14/08, MARK CALLAGHAN <[EMAIL PROTECTED]> wrote: >> >> This is in a future patch. We have implemented global transactions IDs >> (http://code.google.com/p/google-mysql-tools/wiki/GlobalTransactionIds) >> and shared the changes with Mas & Lars. We will publish the patch >> soon. Hopefully we will have per-event checksums in that patch too. > > Checksums for events in the binlog? Yay! > > I've tried to convince the MySQL folks that it's necessary but they want > "proof" that it's a problem in the real world. I've never understood why > they don't understand that network hardware sometimes goes flakey and flips > bits on you and that annoys MySQL slaves. It's more than a theoretical > problem.
They know it is a problem in the real world now and I think checksums are on the list of things they will implement. There was a bug in the MySQL master code that caused corrupt events to be sent to slaves when using flaky network connections. It took them more than a year to fix it and by the end, it was a really big deal for them. > > Jeremy > > > -- Mark Callaghan [EMAIL PROTECTED] _______________________________________________ Mailing list: https://launchpad.net/~drizzle-discuss Post to : drizzle-discuss@lists.launchpad.net Unsubscribe : https://launchpad.net/~drizzle-discuss More help : https://help.launchpad.net/ListHelp
