Andreas Kretschmer <akretschmer(at)spamfence(dot)net> writes:
> No question, this is a silly query, but the problem is the 2nd query: it
> is obviously not possible for the planner to put the where-condition
> into the subquery.
Well, yeah: it might change the results of the window functions.
I see no bug here. Your second query asks for a much more complicated
computation, it's not surprising it takes longer.
regards, tom lane
In response to
pgsql-performance by date
|Next:||From: Andreas Kretschmer||Date: 2010-01-14 17:30:25|
|Subject: Re: bad execution plan for subselects containingwindowing-function|
|Previous:||From: Andreas Kretschmer||Date: 2010-01-14 17:03:18|
|Subject: bad execution plan for subselects containing windowing-function|