Re: Rebuild table
From: Radoulov, Dimitre <cichomitiko_at_gmail.com>
Date: Wed, 22 Jun 2011 09:52:08 +0200
Message-ID: <4E019F28.7090804_at_gmail.com>
On 22/06/2011 03:47, Michael Rosenblum wrote:
> Extra 2c. I would suggest as a possible solution already mentioned
> ALTER TABLE ... MOVE, but fired twice:
> - first, move everything to completely different tablespace
> - second, move it back where the data belongs
> - rebuild indexes
> To be fair, I don't exactly remember the real case, but sometimes ago
> (Oracle 10g R2 on Win32) this double-jump solved my space allocation
> problem, while single operation was not not giving me what I wanted.
Date: Wed, 22 Jun 2011 09:52:08 +0200
Message-ID: <4E019F28.7090804_at_gmail.com>
On 22/06/2011 03:47, Michael Rosenblum wrote:
> Extra 2c. I would suggest as a possible solution already mentioned
> ALTER TABLE ... MOVE, but fired twice:
> - first, move everything to completely different tablespace
> - second, move it back where the data belongs
> - rebuild indexes
> To be fair, I don't exactly remember the real case, but sometimes ago
> (Oracle 10g R2 on Win32) this double-jump solved my space allocation
> problem, while single operation was not not giving me what I wanted.
Yes,
this happens when the default initial/next (for dmt) or uniform (for
lmt) size is big.
Bear in mind that in that case although the space allocated could be large,
the HWM is reset (and a full table scan, for example, actually reads
only up to the HWM).
Dimitre
-- http://www.freelists.org/webpage/oracle-lReceived on Wed Jun 22 2011 - 02:52:08 CDT