Re: Too much redo! Help!

Another option is to take the database out of archive log mode during the update and put it back in afterwards. Of course you'll have a break in your archived logs then and should do a backup immediately after the update completes. To cover yourself.

On 8 Jan 1999 01:33:48 GMT, markp28665_at_aol.com (MarkP28665) wrote:

>From: "Kenny Gump" >>
>If you are recreating the table each time then in your create table statement
>use the unrecoverable keyword option. <<
>
>Original post states version is 7.1.6 so this is not an option, but it a good
>idea once an upgrade takes place and it gave me a second idea. Wheather the
>table is being re-created or just re-loaded with data the original data has to
>be deleted so Tabassam if you are dropping the table or doing a delete of all
>the existing rows use the truncate command to speed the process by reducing the
>amount of rbs and also redo used. It is not much but at least we thought to
>ask. Good luck.
>
>Tabassam Kayani wrote in message >>>...
>We have an overnight procedure which creates a large denormalised table by
>joining a number of existing tables. Unfortunately, more often than not, it
>fails to complete due to the large amount of redo generated (archived) and
>running out of disk space as a consequence.
>
>Bearing in mind that we are running under 7.1.6, can someone please suggest
>how we can keep the redo generated to an absolute minimum ? <<<
>
>
>Mark Powell -- Oracle 7 Certified DBA
>- The only advice that counts is the advice that you follow so follow your own
>advice -

--
Chuck Hamilton
chuckh_at_safeaccess.net

Never share a foxhole with anyone braver than yourself! Received on Mon Jan 11 1999 - 09:34:26 CST