Re: NOLOGGING recovery errors
- From: Hemant K Chitale <hemantkchitale@xxxxxxxxx>
- To: lyallbarbour@xxxxxxxxxxxxxxx
- Date: Wed, 8 Dec 2010 00:03:22 +0800
Just because "NOLOGGING got you into this mess" is no reason to NOT use
NOLOGGING.
If those Indexes are large are you going to spend all that I/O and time to
generate Redo ?
The proper thing to do is REBUILD .. NOLOGGING and then take a backup --
ie ensure that the backup taken after the REBUILD completes is used.
Hemant
On Tue, Dec 7, 2010 at 11:20 PM, <lyallbarbour@xxxxxxxxxxxxxxx> wrote:
> Isn't NOLOGGING how i got into this mess in the first place? And i think
> Grid did use PARALLEL for some, i'll have to check that.
> Lyall
>
>
> -----Original Message-----
> From: Hemant K Chitale <hemantkchitale@xxxxxxxxx>
> To: lyallbarbour@xxxxxxxxxxxxxxx
> Cc: oracle-l@xxxxxxxxxxxxx
> Sent: Mon, Dec 6, 2010 7:29 pm
> Subject: Re: NOLOGGING recovery errors
>
> An ALTER INDEX ..... REBUILD attempts to re-read the existing index
> *unless it is UNUSABLE*.
> So, for indexes that are not enforcing PK constraints, do an ALTER INDEX
> ..... UNUSABLE before the REBUILD.
> For indexes that are for PK constraints, disable the constraints before
> setting them UNUSABLE.
> Needless to say, I suggest NOLOGGING for the REBUILD. If you also use
> PARALLEL then set them to NOPARALLEL after the REBUILD.
> Enable the constraints after the REBUILD.
> Hemant K Chitale
> http://hemantoracledba.blogspot.com
> sent from my smartphone
>
>
>
Other related posts:
