Brian May <[EMAIL PROTECTED]> writes: > My first random guess would be that the packets aren't getting through, > perhaps as a result of a firewall. Either that or the server simply > isn't listening on the right port.
No, they are actually communicating. For the timeout with a 0.6 master and 0.7 slave, Johan Danielsson points out `You can work around this with --time-lost=1year (assuming you update your database more often than that)'. It might be worth noting that. It turns out I mistook similar symptoms with the 0.7 master for the same cause. The slaves die if the master was restarted, which is the sort of thing I thought changes in 0.7 were meant to address. I've asked whether or not that's a bug there, but otherwise it's running OK. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
