Re: V$SQL Executions
From: Jay Hostetter <hostetter.jay_at_gmail.com>
Date: Mon, 19 Sep 2011 15:36:42 -0400
Message-ID: <CAD7fdYt7tp48zsRE0__A3O6dyxAz8181_Qf1oJsa-44hXDWHjA_at_mail.gmail.com>
I did look at those fields as well. It turns out that the application is building queries on the fly and it validates the query as the statement is being built. So the base query gets validated, then the user adds the where clause, then the query is actually executed. Hence, the final query doesn't match the one that was originally parsed. So I have a bunch of parsed, but unexecuted queries. Seems like a waste of CPU time. Thank you,
Jay
Date: Mon, 19 Sep 2011 15:36:42 -0400
Message-ID: <CAD7fdYt7tp48zsRE0__A3O6dyxAz8181_Qf1oJsa-44hXDWHjA_at_mail.gmail.com>
I did look at those fields as well. It turns out that the application is building queries on the fly and it validates the query as the statement is being built. So the base query gets validated, then the user adds the where clause, then the query is actually executed. Hence, the final query doesn't match the one that was originally parsed. So I have a bunch of parsed, but unexecuted queries. Seems like a waste of CPU time. Thank you,
Jay
On Mon, Sep 19, 2011 at 11:48 AM, Hemant K Chitale <hemantkchitale_at_gmail.com
> wrote:
> Also look at cpu_time, disk_reads, buffer_gets etc. SQL executions that
> began but were interrupted / cancelled / killed.
>
> Hemant K Chitale
> http://hemantoracledba.blogspot.com
> http://hemantscribbles.blogspot.com
>
> sent from my smartphone
>
>
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Sep 19 2011 - 14:36:42 CDT