Problems with ordering (can't force query planner to use an index)

Hey,

I have a table that links content together and it currently holds
about 17 mio records. Typical query is a join with a content table and
link table:

noovo-new=# explain analyze SELECT "core_accessor"."id",
"core_accessor"."content_type_id",
"core_accessor"."object_id", "core_accessor"."ordering",
"core_accessor"."label", "core_accessor"."date_posted",
"core_accessor"."publish_state", "core_accessor"."nooximity_old",
"core_accessor"."rising", "core_accessor"."nooximity",
"core_accessor"."nooximity_old_date_posted",
"core_accessor"."nooximity_date_posted", "core_accessor"."user_id",
"core_accessor"."slot_id", "core_accessor"."slot_type_id",
"core_accessor"."role", "core_base"."object_id",
"core_base"."content_type_id", "core_base"."abstract",
"core_base"."abstract_title", "core_base"."image",
"core_base"."date_posted", "core_base"."date_modified",
"core_base"."date_expires", "core_base"."publish_state",
"core_base"."location", "core_base"."location_x",
"core_base"."location_y", "core_base"."raw", "core_base"."author_id",
"core_base"."excerpt", "core_base"."state_id",
"core_base"."country_id", "core_base"."language",
"core_base"."_identifier",
"core_base"."slot_url", "core_base"."source_id",
"core_base"."source_content_type_id", "core_base"."source_type",
"core_base"."source_value", "core_base"."source_title",
"core_base"."direct_to_source", "core_base"."comment_count",
"core_base"."public" FROM "core_accessor" INNER JOIN core_base AS
core_base ON core_base.content_type_id =
core_accessor.content_type_id AND core_base.object_id =
core_accessor.object_id WHERE (("core_accessor"."slot_type_id" = 119
AND "core_accessor"."slot_id" = 472 AND "core_accessor"."label" = E''
AND "core_accessor"."publish_state" >= 60 AND
"core_accessor"."role" IN (0) AND "core_accessor"."user_id" = 0))
order by core_accessor.date_posted, core_accessor.nooximity LIMIT 5
;

QUERY PLAN
---------------------------------------------------------------------------------------------------------------------------------------------------------------------
Limit (cost=31930.65..31930.66 rows=5 width=860) (actual
time=711.924..711.927 rows=5 loops=1)
-> Sort (cost=31930.65..31937.80 rows=2861 width=860) (actual
time=711.923..711.923 rows=5 loops=1)
Sort Key: core_accessor.date_posted, core_accessor.nooximity
Sort Method: top-N heapsort Memory: 31kB
-> Nested Loop (cost=0.00..31883.13 rows=2861 width=860)
(actual time=0.089..543.497 rows=68505 loops=1)
-> Index Scan using core_accessor_fresh_idx on
core_accessor (cost=0.00..5460.07 rows=2970 width=92) (actual
time=0.068..54.921 rows=69312 loops=1)
Index Cond: ((slot_id = 472) AND (slot_type_id =
119) AND (label = ''::text) AND (user_id = 0) AND (role = 0) AND
(publish_state >= 60))
-> Index Scan using core_base_pkey on core_base
(cost=0.00..8.88 rows=1 width=768) (actual time=0.004..0.005 rows=1
loops=69312)
Index Cond: ((core_base.object_id =
core_accessor.object_id) AND (core_base.content_type_id =
core_accessor.content_type_id))
Total runtime: 712.031 ms
(10 rows)

noovo-new=# select * from pg_stat_user_tables where relname='core_accessor';
relid | schemaname | relname | seq_scan | seq_tup_read |
idx_scan | idx_tup_fetch | n_tup_ins | n_tup_upd | n_tup_del |
n_tup_hot_upd | n_live_tup | n_dead_tup | last_vacuum
| last_autovacuum | last_analyze | last_autoanalyze
-------+------------+---------------+----------+--------------+----------+---------------+-----------+-----------+-----------+---------------+------------+------------+-------------------------------+-----------------+-------------------------------+------------------
51159 | public | core_accessor | 58 | 749773516 |
13785608 | 149165183 | 9566 | 548 | 347 |
206 | 17144303 | 251 | 2009-03-03 07:02:19.733778-06 |
| 2009-03-03 06:17:47.784268-06 |
(1 row)

noovo-new=# \d+ core_accessor;
Table "public.core_accessor"
Column | Type |
Modifiers | Description
---------------------------+--------------------------+------------------------------------------------------------+-------------
id | bigint | not null
default nextval('core_accessor_id_seq'::regclass) |
flavor | character varying(32) |
|
content_type_id | integer | not null
|
object_id | integer | not null
|
publish_state | smallint | not null
|
date_posted | timestamp with time zone | not null
|
user_id | integer |
|
slot_id | integer |
|
slot_type_id | integer |
|
role | smallint |
|
ordering | integer |
|
author_id | integer |
|
nooximity_old | double precision | default 0.0
|
rising | double precision | default 0.0
|
label | text |
|
nooximity | double precision | not null
default 1.0 |
nooximity_old_date_posted | timestamp with time zone |
|
nooximity_date_posted | timestamp with time zone |
|
Indexes:
"portal_metainfo_pkey" PRIMARY KEY, btree (id)
"portal_metainfo_unique_constr" UNIQUE, btree (content_type_id,
object_id, user_id, slot_id, slot_type_id, role, label) CLUSTER
"core_accessor_date_idx" btree (date_posted, nooximity)
"core_accessor_dated_idx" btree (slot_id, slot_type_id, label,
user_id, role, publish_state, date_posted, nooximity)
"core_accessor_fresh_idx" btree (slot_id, slot_type_id, label,
user_id, role, publish_state)
"core_accessor_popularity_idx" btree (nooximity, date_posted)
Check constraints:
"portal_metainfo_object_id_check" CHECK (object_id >= 0)
"portal_metainfo_owner_id_check" CHECK (slot_id >= 0)
Foreign-key constraints:
"portal_metainfo_accessor_id_fkey" FOREIGN KEY (user_id)
REFERENCES auth_user(id) DEFERRABLE INITIALLY DEFERRED
"portal_metainfo_content_type_id_fkey" FOREIGN KEY
(content_type_id) REFERENCES django_content_type(id) DEFERRABLE
INITIALLY DEFERRED
"portal_metainfo_owner_type_id_fkey" FOREIGN KEY (slot_type_id)
REFERENCES django_content_type(id) DEFERRABLE INITIALLY DEFERRED
Has OIDs: no

As far as I understand the explain, it fetches 68505 rows, matches
them with core_base and then tries to sort them? AFAIK it would
probably be much more effective to just find the records in accessor
via core_accessor_dated_idx and then lookup the core_base table? But
for some reason it doesn't want to?

I ran analyze, vacuum and reindex but nothing helped. Queries just eat
all the I/O and block. There is a huge difference between cached and
non-cached queries, like 50.000 to 50 ms.

Help! :)

Thanks, Sebastjan