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: LMT advice

Re: LMT advice

From: Brian Peasland <dba_at_remove_spam.peasland.com>
Date: Mon, 29 Sep 2003 14:30:06 GMT
Message-ID: <3F7841EE.AEC039CA@remove_spam.peasland.com>

> How it got there? As I mentioned, even though 1MB or 64MB extents are
> allocated, under the hood, everything in autoallocated tablespace is
> actually allocated as 64K subextents. Your odd-sized extents are whole
> multiples of 64K, so I'm not surprised to see them there.

I'm not surprised to see them there either.  

> It also goes to demonstrate that the 64K sub-extents model does a perfectly
> fine job of avoiding fragmentation (ie, pockets of free space that nothing
> can make use of. Drop an 8M extent, and you have actually created 128
> pieces of free space, which a future 1M extent will slot into easily, and
> which a future 8M extent will then be able to take 7M from, and find 1M
> elsewhere.

Exactly! From what I've been able to discern in my tests is just as you describe. Autoallocate wants to allocate an extent of a certain size, but that does not mean that the algorithm won't actually allocate more than one extent totalling the size it initially wanted to allocate.

Cheers,
Brian

-- 
===================================================================

Brian Peasland
dba_at_remove_spam.peasland.com

Remove the "remove_spam." from the email address to email me.


"I can give it to you cheap, quick, and good. Now pick two out of
 the three"
Received on Mon Sep 29 2003 - 09:30:06 CDT

Original text of this message

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