re: find dirty blocks/records; 9.2.0.7
- From: Cosmin Ioan <cosmini@xxxxxxxxxxxxxxx>
- To: oracle-l@xxxxxxxxxxxxx
- Date: Mon, 14 May 2007 07:30:59 -0700 (PDT)
hi all,
I'm working on a (telecom) high transactional system where many individual
records could be in a dirty state (updated/deleted but not commited until a
later time, possibly minutes or more). It is possible that on occasion, we
need to run batch processes on bulk records, say millions or more.
While we're trying to limit this large processing at off-peak hours -- to
limit contention with production, I was wondering if there is any way to
identify the dirty blocks/records so that we can re-visit them later for
processing. Unfortunately, there is no one process that may update these
individual records and as such a flag can be set, so I need to go to the
database level, to the dictionary and get my dirty records.
Theoretically, it seems to me that perhaps the fastest way to do this is via
BULK/FORALL statements as this technique will allow for individual records
exception (due to locking or others), rather than simple SQL set based updates
which could likely generate row and even table level locking? Am I going
about it the right way?
thx much,
Cosmin
Other related posts:
- » re: find dirty blocks/records; 9.2.0.7
