Indexing problem with OFFSET LIMIT

["Oliver Weichhold" <oliver@xxxxxxxxxxxxx>]

Hello

I have problem in my applications and don't know how to fix it.

This is the table and one of the indexes:

CREATE TABLE foo
(
  id serial NOT NULL,
  foo_name character varying(100),
  realm_id integer

  ... and about 50 other columns
)

CREATE INDEX idx_foo_name_realm
  ON foo
  USING btree
  (realm_id, foo_name);

Table foo contains about 8 Million Rows. 


The problem:

Consider this query:

SELECT * FROM foo WHERE realm_id = 228 order by foo_name LIMIT 200 OFFSET 15000

And it's execution plan:

"Limit  (cost=57527.13..58294.16 rows=200 width=575) (actual time=182.302..184.971 rows=200 loops=1)"
"  ->  Index Scan using idx_foo_name_realm on foo  (cost=0.00..62159.98 rows=16208 width=575) (actual time=0.085..166.861 rows=15200 loops=1)"
"        Index Cond: (realm_id = 228)"
"Total runtime: 185.591 ms"


And now look at this:

SELECT * FROM foo WHERE realm_id = 228 order by foo_name LIMIT 200 OFFSET 15999

"Limit  (cost=59601.92..59602.42 rows=200 width=575) (actual time=1069.759..1072.310 rows=200 loops=1)"
"  ->  Sort  (cost=59561.92..59602.44 rows=16208 width=575) (actual time=929.948..1052.620 rows=16199 loops=1)"
"        Sort Key: foo_name"
"        Sort Method:  external merge  Disk: 8984kB"
"        ->  Bitmap Heap Scan on foo  (cost=306.69..54270.62 rows=16208 width=575) (actual time=9.612..235.902 rows=21788 loops=1)"
"              Recheck Cond: (realm_id = 228)"
"              ->  Bitmap Index Scan on foo_realm_id  (cost=0.00..302.64 rows=16208 width=0) (actual time=8.733..8.733 rows=21810 loops=1)"
"                    Index Cond: (realm_id = 228)"
"Total runtime: 1084.706 ms"

Execution time increases tenfold because postgres stopped using the index.

Can anybody explain to me what's going on and what can be done? Is this a memory problem?