Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102
| От | Vladimir Sitnikov |
|---|---|
| Тема | Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102 |
| Дата | |
| Msg-id | CAB=Je-EqbGdaNauAgpzULSeEBQKw5c-b4GVp9C7k5W4cJGpUXg@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: Re: 9.4-1207 behaves differently with server side prepared statements compared to 9.2-1102 (Dave Cramer <pg@fastcrypt.com>) |
| Список | pgsql-jdbc |
Thomas, as far as I see, you have offending queries and bind values for them, don't you? Can you please perform the following dance in psql or pgadmin? PREPARE prep1207(number,number,number,number,number,number) as -- <-- data types for the bind variables select .. $1, ... where id=$2; -- <-- your query here EXPLAIN ANALYZE EXECUTE prep1207(1,2,3,42,100500); -- <-- values for bind variables EXPLAIN ANALYZE EXECUTE prep1207(1,2,3,42,100500); -- <-- values for bind variables EXPLAIN ANALYZE EXECUTE prep1207(1,2,3,42,100500); -- <-- values for bind variables EXPLAIN ANALYZE EXECUTE prep1207(1,2,3,42,100500); -- <-- values for bind variables EXPLAIN ANALYZE EXECUTE prep1207(1,2,3,42,100500); -- <-- values for bind variables Ideally, it should reproduce the issue. In other words, 5th or 6th execution should flip to bad plan. Then share the plan/query. If table names cannot be made public, you might want try http://explain.depesz.com/ to anonymize the plan. Vladimir
В списке pgsql-jdbc по дате отправления: