Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Clarifying something in the online docs

Re: Clarifying something in the online docs

From: BD <robert.drea_at_gmail.com>
Date: 27 Mar 2007 15:15:58 -0700
Message-ID: <1175033758.475212.146400@y80g2000hsf.googlegroups.com>


On Mar 27, 2:51 pm, Richard Piasecki <usen..._at_ogoent.com> wrote:
> Is there any reason why you don't want to use a refresh group for these mviews?

Historically (we're moving from an ODS export/import scenario), automation of the replication has been table-driven; the list of tables to be replicated were governed by a driving table, to which a record would be added for each table to be replicated.

The requirement of maintaining a refresh group, and ensuring that new tables get included in it through the use of the ADD procedure, would likely require as much automation coding (if not more) than the code required to generate the script for the individual refresh statements... and since that process has already been done, the additional benefit that a refresh group would have to my situation is unclear: there's no need for me to ensure the point-in-time consistency that a refresh group allows for, and I only have 'the one group' - so there's no benefit of creating multiple groups for multiple schedules, etc.

If there were a 'generic', schema-wide, single refresh procedure that didn't require the maintenance of the group, I'd probably go for it.

In fact, there is a benefit of me being able to report on how long each of these mvs is taking to refresh - so doing them individually with TIMING set to ON is helpful for maintaining metrics on the volume of changes being made to the source tables. Received on Tue Mar 27 2007 - 17:15:58 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US