Alvaro Herrera <alvhe...@2ndquadrant.com> writes: >> - will the table have to be rewritten? the indexes? > > Please give my DDL deparsing patch a look. There is a portion there > about deparsing ALTER TABLE specifically; what it does is save a list of > subcommands, and for each of them we either report the OID of the object > affected (for example in ADD CONSTRAINT), or a column number (for ALTER > COLUMN RENAME, say). It sounds like you would like to have some extra > details returned: for instance the "does the whole of it require a table > rewrite" bit. It sounds like it can be trivially returned in the JSON
Some years ago when working on the Event Trigger framework we did mention providing some interesting events, such as a TableRewrite Event. In between what you're saying here and what Harold and Peter Geoghegan are mentionning (basically that dealing with table rewrites is 90% of the need for them), it could be that the best way to have at it would be to add that Event in the Event Trigger mechanism. We could also add an AccessExclusiveLock Event that would fire just before actually taking the lock, allowing people to RAISE EXCEPTION in that case, or to maybe just do the LOCK … NOWAIT themselves in the trigger. For the locking parts, best would be to do the LOCK … NOWAIT dance for all the tables touched by the DDL migration script. The Event Trigger approach will not solve that, unfortunately. Regards, -- Dimitri Fontaine 06 63 07 10 78 http://2ndQuadrant.fr PostgreSQL : Expertise, Formation et Support -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
