I'm trying to access a table with about 120M rows. It's a vertical version
of a table with 360 or so columns. The new columns are: original item col,
original item row, and the value.
I created an index:
CREATE INDEX idx on table (col, row)
however, selects are still very slow. It seems it still needs a sequential
scan:
EXPLAIN SELECT * FROM table WHERE col=1 AND row=10;
QUERY PLAN
------------------------------------------------------------------------------
Seq Scan on table (cost=100000000.00..102612533.00 rows=1 width=14)
Filter: ((col = 1) AND ("row" = 10))
What am I doing wrong?
--
Dan Pelleg