Re: [PERFORM] Inefficient query plan

2010-08-24 Thread Jann Röder
So that took a while... I'm currently running ANALYZE on the PaperReferences table again (the one where I changed the data type). The plan however is still the same: "Hash Join (cost=280.88..24330800.08 rows=670602240 width=16)" " Hash Cond: (r.itemid = p.itemid)" " -> Seq Scan on paperreferen

Re: [PERFORM] Inefficient query plan

2010-08-24 Thread Jann Röder
Thanks everyone, the problem just solved itself. After the ANALYZE had finished, postgres started doing what I wanted it to do all along: EXPLAIN SELECT p.ItemID FROM Papers AS p, PaperReferences AS r WHERE p.itemID = r.ItemID AND p.issueID = 'A1983PW823'; "Nested Loop (cost=0.00..4515980.97 rows

Re: [PERFORM] Are Indices automatically generated for primary keys?

2010-08-24 Thread Guillaume Lelarge
Le 18/08/2010 17:23, Thom Brown a écrit : > On 18 August 2010 17:06, Justin Graf wrote: >> On 8/18/2010 9:15 AM, Clemens Eisserer wrote: >>> Hi, >>> >>> they are generated automatically. >>> Thanks depesz! >>> The reason why I asked was because pgAdmin doesn't display the >>> automatical