Re: WRI$_OPTSTAT_TAB_HISTORY not showing table stats on certain days: why?
- From: fmhabash <fmhabash@xxxxxxxxx>
- To: rjamya <rjamya@xxxxxxxxx>
- Date: Tue, 10 Jan 2012 14:45:53 -0500
- no truncate
- It did not run out of time as the weekend windows lasts for 20 hrs
(and auto gather wakes up every 4 hrs) and the job elapsed time on both
days was less than an hour.
- I can see why if the table persisted at 0 rows (at least when the job
invoked), but it did the same on previous consecutive days and it still
got analyzed.
On 1/10/2012 1:51 PM, rjamya wrote:
> I guess there are two main reasons ...
> 1. maybe auto job decided that this table didnt have sufficient
> changes to warrant the new stats collection (which you already mentioned)
> 2. or it ran out of time (job window).
> does this table gets truncated? if so, then stats will still be
> collected even if rowcount is 0.
> It could be either reason or could be some other "undocumented" reason ..
> Raj
> On Tue, Jan 10, 2012 at 1:36 PM, fmhabash <fmhabash@xxxxxxxxx
> <mailto:fmhabash@xxxxxxxxx>> wrote:
>
> Hi all ..
> I'm trying to understand why WRI$_OPTSTAT_TAB_HISTORY is not showing
> this table to be analyzed on 7th and 8th (weekend days). Take into
> consideration ...
> - I confirmed auto stats job did run on these 2 days (starting at
> 6am).
> - If the explanation that it did not qualify because of the % stale,
> then why it did on previous days when rowcnt was 0.
>
--
//www.freelists.org/webpage/oracle-l
Other related posts:
