From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | demurcia(at)ockham-solutions(dot)fr |
Cc: | pgsql-bugs(at)postgresql(dot)org |
Subject: | Re: BUG #7593: Unexpected query result combining or and in operators |
Date: | 2012-10-10 17:17:32 |
Message-ID: | 1197.1349889452@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-bugs |
demurcia(at)ockham-solutions(dot)fr writes:
> After a little while (probably after the analyze), the query plan is changed
> and only 2 results are returned !
> "Merge Join (cost=33.31..39.69 rows=1 width=24)"
> " Merge Cond: (i.numero_id = n.numero_id)"
> ...
> " -> Index Only Scan using numero_pkey on numero n (cost=0.00..12.57
> rows=2 width=8)"
> " Index Cond: (numero_id = ANY ('{25393,25384}'::bigint[]))"
Yeah, there's a known bug in 9.2 with mergejoins whose inner side is an
indexscan with an =ANY condition. There's a fix here, if you're able to
apply a patch:
Otherwise, the fix will be in 9.2.2.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | hikkis21c | 2012-10-11 09:00:25 | BUG #7595: terminate process in an unusual way |
Previous Message | max | 2012-10-10 16:02:05 | BUG #7594: "tuple concurrently updated" error on concurrent GRANT ON DATABASE statements |