RE: Insert contention on RAC
From: Tim Gorman <tim_at_evdbt.com>
Date: Wed, 25 Jun 2008 17:09:49 -0600
Message-ID: <20080625170949.qiht3foqo0gsckwg@webmail.evdbt.com>
Date: Wed, 25 Jun 2008 17:09:49 -0600
Message-ID: <20080625170949.qiht3foqo0gsckwg@webmail.evdbt.com>
Quoting "Mark W. Farnham" <mwf_at_rsiz.com>:
> (A Gorman is where you totally avoid contention with an existing big
> partitioned table being affected by dataloads by doing a partition exchange
> after the data is loaded. So this is a modified Gorman, in that you'll be
> exchanging in the n tables to something that was formerly empty.)
>
> (If Tim doesn't like me calling that a Gorman, I'll withdraw it.)
...well, it's better than the previous definition of a "Gorman", which involves a bone-headed move resulting in serious injury that makes people cringe on the replay...
...and I don't even want to think about the previous definition of a "modified Gorman", except that it makes dogs start howling...
-- http://www.freelists.org/webpage/oracle-lReceived on Wed Jun 25 2008 - 18:09:49 CDT