From: | Stefan Keller <sfkeller(at)gmail(dot)com> |
---|---|
To: | pgsql-general List <pgsql-general(at)postgresql(dot)org> |
Subject: | hstore equality-index performance question |
Date: | 2010-03-28 22:57:20 |
Message-ID: | 25bc040b1003281557l72df790csd818d1b258043744@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Documentation at "F.13.3. Indexes" says that "hstore has index support
for @> and ? operators..."
=> Therefore no index does support equality-indexes?
If so, then I suppose that following (potentially slow) query
which contains an equality test for all keys 'a' and returns all values...
SELECT id, (kvp->'a') FROM mytable;
... can be accelerated nevertheless by adding following where clause:
SELECT id, (kvp->'a') FROM mytable WHERE kvp ? 'a';
=> Is this correct?
-S.
-- Little test database
CREATE TABLE mytable (
id serial PRIMARY KEY,
kvp HSTORE
);
CREATE INDEX mytable_kvp_idx ON mytable USING GIN(kvp);
INSERT INTO mytable (kvp) VALUES ('a=>x, b=>y');
INSERT INTO mytable (kvp) VALUES ('a=>y, c=>z, d=>a');
From | Date | Subject | |
---|---|---|---|
Next Message | Filip Rembiałkowski | 2010-03-28 23:37:20 | Re: optimizing import of large CSV file into partitioned table? |
Previous Message | David Kerr | 2010-03-28 22:25:17 | Re: Connection Pooling |