Question

I have index on a column, let’s say ID (bigint). If I have a query with something like this:

SELECT * 
  FROM table 
 WHERE id = 12345

…it will use index. But when I’m using query like…

SELECT * 
  FROM table 
 WHERE id >= 12345 
   AND id <= 12366

It use sequential scan, which is very slow. Can I force using the ID index?

Answer 1

It should use the index if the index type is btree and select doesn’t fetch more then 30% of all record count (is it true in postgresql as well?) @scott-marlowe says that “..for PostgreSQL the switchover point comes much earlier, somewhere in the 1 to 10% range where it’s cheaper to do a sequential scan..”.

Try calling REINDEX action maybe?