Tom, The Analyze did in fact fix the issue. Thanks.
--sean On 9/27/04 11:54 PM, "Sean Shanny" <[EMAIL PROTECTED]> wrote: > Tom, > > We have been running pg_autovacuum on this entire DB so I did not even > consider that. I am running an analyze verbose now. > > We should see about 82mm rows that will match the Filter: ((date_key >= 610) > AND (date_key <= 631)) > > I'll update in an hour or so. > > --sean > > > On 9/27/04 11:49 PM, "Tom Lane" <[EMAIL PROTECTED]> wrote: > >> Sean Shanny <[EMAIL PROTECTED]> writes: >>> -> Seq Scan on f_pageviews t1 (cost=0.00..11762857.88 >>> rows=1 width=8) >>> Filter: ((date_key >= 610) AND (date_key <= 631)) >> >> How many rows are actually going to match that filter condition? (The >> symptoms seem to indicate that the answer is "a whole lot", not "1".) >> >> I speculate that you're overdue for an ANALYZE on this table, and that >> the planner thinks this scan is going to yield no rows because the >> stats it has say there are no rows with date_key >= 610. >> >> regards, tom lane > > > > ---------------------------(end of broadcast)--------------------------- > TIP 6: Have you searched our list archives? > > http://archives.postgresql.org ---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
