On 4/26/07, Andrew Sullivan <[EMAIL PROTECTED]> wrote: > On Thu, Apr 26, 2007 at 03:03:55PM +0200, Paul Slootman wrote: > > > Hmm, that sounds a bit worrying (the nasty bugs, that is)... Slony 1.0.5 > > has worked pretty well for me up to now, but now the application people > > are starting to ask for postgresql 8 as that should help performance, they > > say. I say they should fix their application, but do they listen.... :-) > > PostgreSQL 8.x series actually does have significant performance > advantages for even correctly-written applications, because 8.0 is [snip]
OK, sounds sensible. > As for Slony, the bug problems have mostly been subtle things. The > point is more generally, though, that it's a critical piece of your > infrastructure, and there is not a long history of experience with > it. Because Debian has such strict rules about upgrade packages > after freeze (a policy of which I generally approve, BTW), you can > find yourself living with a possibly broken package for a long time. > So even though generally I'm not a fan of deviating from the > standard stable package, if you're going to use a packaged slony, I > urge you to be ready to backport a new version. Well, I'll have to backport whatever new version of slony I choose to my postgresql 7.4 version anyway in order to upgrade without significant downtime, so it'll happen anyway. That said, building slony 1.2.9 for pg 7.4 seemed to go well just now... Now to schedule a window for upgrading slony on my existing cluster. Thanks for your comments. Paul Slootman _______________________________________________ Slony1-general mailing list [email protected] http://gborg.postgresql.org/mailman/listinfo/slony1-general
