Wed Aug 10 19:09:52 PDT 2005
- Previous message: [Slony1-general] Slony >> Multi Master Replication???
- Next message: [Slony1-general] win32 service code
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Wed, Aug 10, 2005 at 11:02:32AM -0400, Jan Wieck wrote: > > First one wins? I thought last update wins is the default for that sort. I'm just reporting what I heard. > Both are insufficient for resolving conflicts on balance type columns > (like an account balance). What would be necessary is the distinction No guff ;-) > There are more complicated race conditions. Think of a 3 node setup. > Node A does an INSERT. Node B replicates that and then does an UPDATE to > that row. How do you guarantee that node C will replicate the INSERT > before replicating the UPDATE? You got me. This is why I hate such systems -- they're too hard to understand, and I always feel like something's going to get undetectably lost. A -- Andrew Sullivan | ajs at crankycanuck.ca In the future this spectacle of the middle classes shocking the avant- garde will probably become the textbook definition of Postmodernism. --Brad Holland
- Previous message: [Slony1-general] Slony >> Multi Master Replication???
- Next message: [Slony1-general] win32 service code
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list