Slightly similarly, we had a case where rebuilding the snapshot log
helped.
A week ago we had noticed that MV refresh's were taking longer and were
occassionally failing
on ORA-1555s. We then redefined the MV to extract a smaller subset
of data . We knew
that this was supposed to be faster. Yet, it wasn't any faster and
the load that the refresh
was causing was still high. We realised that the snapshot logs on
the base tables had
grown inordinately large {as a result of the ORA-1555s in the
refresh's}. We dropped
and recreated the snapshot logs. Thereafter, MV refresh's were also much
faster.
Hemant
At 12:33 AM Thursday, rjamya wrote:
Just for AQ
this script on metalin is also good.
Raj
On 10/26/05, Stephane Faroult
<sfaroult@roughsea.com>
wrote:
- There is something else that may benefit, namely tables that are
used
- for queues and that had their high-water mark bumped much too high at
a
- given moment. But copying their contents elsewhere, truncating
and
- re-inserting the rows definitely don't require reorganizing the full
- database ... I agree with the general feeling: waste of time.
Great
- opportunity to say "told you so".
- Stephane Faroult
Hemant K Chitale
http://web.singnet.com.sg/~hkchital
--
http://www.freelists.org/webpage/oracle-l
Received on Thu Oct 27 2005 - 10:12:45 CDT