Sort
От | Nathan Barnett |
---|---|
Тема | Sort |
Дата | |
Msg-id | 71975481CD04D4118E57004033A2596E0DF93E@ip205.82.136.216.in-addr.arpa обсуждение исходный текст |
Список | pgsql-general |
I'm currently using v7.0.2 of PostgreSQL. I have a query that performs a group by on three columns. The EXPLAIN of the query is as follows: Aggregate (cost=4116.05..4125.47 rows=94 width=24) -> Group (cost=4116.05..4123.12 rows=942 width=24) -> Sort (cost=4116.05..4116.05 rows=942 width=24) -> Nested Loop (cost=0.00..4069.52 rows=942 width=24) -> Seq Scan on click (cost=0.00..15.42 rows=942 width=8) -> Index Scan using impression_pkey on impression (cost=0.00..4.29 rows=1 width=16) I need to speed up this query. I have already created an index on the three columns query_idx(columna, columnb,columnc). Is there some other index that I could add which would speed up the query. There is a high frequency in the table of each group. Each group probably makes up 10% of the table. Does this force a sequence scan when sorting and grouping? Basically just looking for suggestions. ---------------- Nathan Barnett
В списке pgsql-general по дате отправления: