| Subject | Re: [Firebird-Architect] "Write Committed" transaction mode |
|---|---|
| Author | Helen Borrie |
| Post date | 2010-11-30T00:06:41Z |
At 11:03 AM 30/11/2010, Jim Starkey wrote:
It seems to me more sensible/obvious than the hairy (IMHO) mechanism behind FOR UPDATE WITH LOCK as implemented in Fb 1.5. Could you describe with just a little more detail (with reference to the current situation with Snapshot and both of the resolution options of ReadCommitted ) the lock conflicts that could occur and the options for resolving them.
Helen
>So, it's a pessimistic-optimistic lock..? ;-)
>In both Falcon and NimbusDB, a select for update (either implicit or
>explicit) in "write committed" mode cause a dummy update similar to an
>Interbase/Firebird deleted record stub, but with a distinctive record
>number indicating "record lock."
It seems to me more sensible/obvious than the hairy (IMHO) mechanism behind FOR UPDATE WITH LOCK as implemented in Fb 1.5. Could you describe with just a little more detail (with reference to the current situation with Snapshot and both of the resolution options of ReadCommitted ) the lock conflicts that could occur and the options for resolving them.
Helen