[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ext4
Subject:    [Ext2-devel] Re: [lkml FW]: ext3 reservation strangeness?!
From:       karl.vogel () seagha ! com
Date:       2004-10-26 20:02:59
Message-ID: m31xfll024.fsf () seagha ! com
[Download RAW message or body]

Mingming Cao <cmm@us.ibm.com> writes:

> 1) ext3_find_goal() randomly(based on the pid) select a goal block for
> the very first block to allocate. It is possible that ext3_find_goal
> generate a bad start for the reservation case-- the bitmap around the
> goal block has a lot of other files already allocated. But for the
> non-reservation case, it picked up a good place to place the file, there
> is no(or only a few) other files nearby. I don't know if we could do
> something here.

I don't know the ext3 code, but I thought that the actual allocation only
happened at the write out moment?! So my understanding was that if I 
copied the 8mb file (which can be cached entirely in memory), that it
would be allocated on disk when it was actually written out and with
enough diskspace available, it should be possible to find a place on
disk that could fit the file without fragments. (that's also why I raised
the commit time, as I thought it caused the allocation to occur faster).

Or is this not the way it works?!


-------------------------------------------------------
This SF.Net email is sponsored by:
Sybase ASE Linux Express Edition - download now for FREE
LinuxWorld Reader's Choice Award Winner for best database on Linux.
http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click
_______________________________________________
Ext2-devel mailing list
Ext2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ext2-devel
[prev in list] [next in list] [prev in thread] [next in thread]