[Top] [All Lists]

Re: Zero filled files

To: Steve Lord <lord@xxxxxxx>
Subject: Re: Zero filled files
From: Andi Kleen <ak@xxxxxxx>
Date: Tue, 13 May 2003 17:46:29 +0200
Cc: Andi Kleen <ak@xxxxxxx>, linux-xfs@xxxxxxxxxxx
In-reply-to: <1052839608.22728.104.camel@xxxxxxxxxxxxxxxxxxxx>
References: <20030512102338.GA3268@xxxxxxxxxxxxxxx> <1052742135.1173.1.camel@xxxxxxxxxxxxxxxxxxxxxxx> <20030513053133.GG10596@xxxxxxxxxxxxx> <1052839608.22728.104.camel@xxxxxxxxxxxxxxxxxxxx>
Sender: linux-xfs-bounce@xxxxxxxxxxx
On Tue, May 13, 2003 at 10:26:49AM -0500, Steve Lord wrote:
> Complete change in approach is the short answer.


> By adding the periodic actvity thread we activate some code which
> looks for an empty AIL and writes out a dummy log record to record
> the new tail of the log.

Does it check that it already wrote the marker before writing again ? 
If not the laptop users will hate you ;)

> There are still windows when zero filled files are possible, as the
> updated inode size can make it out to disk in a transaction before
> all the extents do. Doing the 100% solution will require some brain
> cells.

And I imagine changing this could add lots of seeking ...
(between the inode areas and the data areas) 

> I can say though, that after sync returns a linux xfs filesystem is
> now on disk to the point where it will look the same after a reboot.

That should fix the number 1 complaint about XFS yes.

Thanks for the explanation,

<Prev in Thread] Current Thread [Next in Thread>