On Mon, Dec 15, 2014 at 11:55:43AM -0800, Paul Theodoropoulos wrote: ... > occasional/rare I mean perhaps a few days will pass, we'll have a > deadlock and the master process will abort/crash and restart. > Obviously we'd prefer not to have the minute or so of downtime as > monit discovers the outage and restarts it. > > Is this a case where just increasing the kid_deadlock_sleep would > help? Is there an upper bounds on the sleep that would tend to make > things wonky? I currently have it set at one second.
No, that setting will not change anything - it just indicates how long Bucardo should wait before trying the kid again when a deadlock is detected. But it won't help prevent them in the first place. The best you can do, other than making your application smart enough to handle deadlocks on its own, is to keep your updates fast, commit often, and make your syncs as long as possible. In other words, make sure everybody gets in and out of the database as fast as possible to prevent such deadlocks. If it becomes a major problem, there are heavier solutions (e.g. table- or advisory- level locks), but for the most part, it's just a natural consequence of living in a MVCC world. :) -- Greg Sabino Mullane [email protected] End Point Corporation PGP Key: 0x14964AC8
signature.asc
Description: Digital signature
_______________________________________________ Bucardo-general mailing list [email protected] https://mail.endcrypt.com/mailman/listinfo/bucardo-general
