Re: WIP: Enhanced ALTER OPERATOR

Поиск
Список
Период
Сортировка
От Andres Freund
Тема Re: WIP: Enhanced ALTER OPERATOR
Дата
Msg-id 20150520185041.GJ27868@alap3.anarazel.de
обсуждение исходный текст
Ответ на Re: WIP: Enhanced ALTER OPERATOR  (Uriy Zhuravlev <u.zhuravlev@postgrespro.ru>)
Ответы Re: WIP: Enhanced ALTER OPERATOR  (Uriy Zhuravlev <u.zhuravlev@postgrespro.ru>)
Список pgsql-hackers
Hi,

On 2015-05-20 12:22:34 +0300, Uriy Zhuravlev wrote:
> On Monday 18 May 2015 10:21:10 you wrote:
> > difficulty of updating existing cached plans
> Could you specify more precisely about some caches we talking about? PREPARE 
> working correctly:
> 
> CREATE TABLE test_ints(i int4);
> CREATE TABLE
> CREATE INDEX idx ON test_ints(i);
> CREATE INDEX
> set enable_bitmapscan=off;
> SET
> set enable_seqscan=off;
> SET
> PREPARE test_plan (int) AS 
>         SELECT * FROM test_ints WHERE $1::int4 > i;
> PREPARE
> EXPLAIN (COSTS OFF)
> EXECUTE test_plan(5);
>                QUERY PLAN               
> ----------------------------------------
>  Index Only Scan using idx on test_ints
>    Index Cond: (i < 5)
> 
> ALTER OPERATOR > (int4, int4) SET COMMUTATOR NONE;
> ALTER OPERATOR
> EXPLAIN (COSTS OFF)
> EXECUTE test_plan(5);
>                QUERY PLAN               
> ----------------------------------------
>  Index Only Scan using idx on test_ints
>    Filter: (5 > i)

Note that this very likely wasn't actually using a prepared plan. Due to
the custom plan infrastructure the first few invocations are going to be
replanned.

Greetings,

Andres Freund



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Tom Lane
Дата:
Сообщение: Re: WIP: Enhanced ALTER OPERATOR
Следующее
От: Bruno Harbulot
Дата:
Сообщение: Re: Problems with question marks in operators (JDBC, ECPG, ...)