Wed Jul 18 09:22:51 PDT 2007
- Previous message: [Slony1-general] Replication and deferrable constraints
- Next message: [Slony1-general] Replication and deferrable constraints
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Wed, Jul 18, 2007 at 08:11:40PM +0400, Dmitry Koterov wrote: > Will those two INSERT go ALWAYS in the same transaction on a subscriber or > not? No. > If not, nobody should use deferrable constraints together with Slony, > because sometimes query sequence correct for master will be incorrect for > slave. Also no. Slony does things in "agreeable order", which is where much of the actual invention in Slony is (this was entirely Jan's idea, and it's a very clever one). That is, it respects the visibility effects of MVCC _without_ having to replay the transactions in exactly the same order as they happened on the origin. The explication of how this is to work is in the "concepts" doc that Jan originally wrote, which Chris already pointed you to. A -- Andrew Sullivan | ajs at crankycanuck.ca However important originality may be in some fields, restraint and adherence to procedure emerge as the more significant virtues in a great many others. --Alain de Botton
- Previous message: [Slony1-general] Replication and deferrable constraints
- Next message: [Slony1-general] Replication and deferrable constraints
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list