Brian Peasland wrote:
> >> Also, did you check the segment type (SEGTYPE) in V$SORT_USAGE to
> >> determine if the temporary segment is really a sort segment? According
> >> to Jonathan Lewis in this Metalink thread
> >> (https://metalink.oracle.com/metalink/plsql/f?p=200:27:1683701837942099652::::p27_id,p27_show_header,p27_show_help:279353.999,1,1),
> >> V$SORT_USAGE will show more than just disk sorts. The sort segment in
> >> V$SORT_USAGE can be temporary tables, open cursors or some temporary
> >> LOBs, in which case the session statistic for 'sorts (disk)' would not
> >> be increased.
> >
> > SEGTYPE is DATA,INDEX,LOB_DATA
> >
>
> If the segment type is not SORT, then the segment being tracked in
> V$SORT_USAGE is not from a sorting operation, therefore, there would not
> be any 'sorts (disk)' to account for. Your data makes sense now.
>
>
make sense, greate answer!
Received on Fri Jun 09 2006 - 11:25:25 CDT
