Re: [HACKERS] Statement-level rollback
От | Tsunakawa, Takayuki |
---|---|
Тема | Re: [HACKERS] Statement-level rollback |
Дата | |
Msg-id | 0A3221C70F24FB45833433255569204D1F6A9D7F@G01JPEXMBYT05 обсуждение исходный текст |
Ответ на | Re: [HACKERS] Statement-level rollback (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-hackers |
From: pgsql-hackers-owner@postgresql.org > [mailto:pgsql-hackers-owner@postgresql.org] On Behalf Of Tom Lane > "Tsunakawa, Takayuki" <tsunakawa.takay@jp.fujitsu.com> writes: > > As I stated here and at the PGConf.ASIA developer meeting last year, > > I'd like to propose statement-level rollback feature. > > I do not really see how this would ever get past the compatibility problems > that forced us to give up on server-side autocommit years ago. Could you tell me more about that problem? What kind of incompatibility would this feature introduce? > If you want to provide a client-side facility for this, perhaps that could > fly. Do you mean a feature of psqlODBC that implicitly issues SAVEPOINT and RELEASE SAVEPOINT for each SQL statement? One reasonI want to implement the feature is to avoid eliminate those round-trips for performance. Or, do you mean a client-sideconnection parameter like "rollback_scope={transaction | statement}?" Yes, I'll implement it for major clientdrivers so that the driver issues "SET SESSION CHARACTERISTICS FOR TRANSACTION ROLLBACK SCOPE {TRANSACTION | STATEMENT}"upon connection. psqlODBC has already a connection parameter, Protocol, for that purpose. Regards Takayuki Tsunakawa
В списке pgsql-hackers по дате отправления: