Re: How to show biggest redo producers

Yup, strangely enough, Log Miner will allow you to analyze the stream of redo contained within redo log files, and that would most certainly be the easiest way to do what you need to do (for a start, you could analyze archived redo logs [assuming you're in archive log mode] using an entirely separate database, and apart from some physical I/O, such an analysis won't impact on your database at all (and if you do the sensible thing and stick archives on entirely their own hard drives, even the I/O won't be an issue for your production database).

However, you might start to need to get a bit clever... analyzing just for "seg_name='EMP'", for example, won't show you all the rollback and index entries that a transaction produces. You'll have to search by serial#, and group all the redo generated by a transaction by that unique identifier.

Regards
HJR

--
----------------------------------------------
Resources for Oracle: http://www.hjrdba.com
===============================


"Tom" <tdwilsonng_at_yahoo.com> wrote in message
news:ae755ef4.0203010539.51494bcd_at_posting.google.com...


> I have a database producing tremendous amounts of redo, and I'd like

> to show which tables' updates are the biggest offenders.  Is there a

> way to get these statistics?  Will log miner work?  Is it the only

> way?

>

> Tom