[Developers] Fwd: disk space usage reporting
David Rideout
dprideout at gmail.com
Fri Oct 20 10:31:29 CDT 2006
Since they still seem relevant to the discussion, I'll forward
messages that got blocked this morning, with:
Delivery to the following recipient failed permanently:
developers at cactuscode.org
Technical details of permanent failure:
PERM_FAILURE: DNS Error: DNS server returned answer with no data
---------- Forwarded message ----------
From: David Rideout <dprideout at gmail.com>
Date: Oct 20, 2006 11:13 AM
Subject: Re: [Developers] disk space usage reporting
To: developers at cactuscode.org
Since you raised the discussion, I'll toss out another thought.
Perhaps, instead of crashing, Cactus could send the user a warning
when it runs low on disk space, then if possible, checkpoint, and if
not, suspend itself. The logic of this could be some function of the
amount of remaining disk space (and its time derivatives?), amount of
disk space required to checkpoint, etc.
Cheers,
David
On 10/20/06, Steve White <steve.white at aei.mpg.de> wrote:
> Let me just toss this idea out:
>
> We have a lot of disk space trouble. It often disturbs users' otherwise
> professional demeanor, and taints their otherwise collaborative atmosphere.
>
> The usual remedies proposed, "get a bigger mass storage device" and
> "stop being an asshole", have so far met with limited success.
>
> Could Cactus help?
>
> I have in mind: if Cactus could calculate and sum up how much data it
> writes in each iteration (and perhaps compare that to remaining storage
> space, and estimate how quickly it will fill it), then our (really very
> polite and cooperative) users could more easily predict when they are
> about to put their elbow in their collegue's tea.
>
> It seems to me that small alterations in the IO utilities could perform
> the collection of file sizes. As to how the rest would work, it would
> require consideration and discussion.
>
> And I'm not proposing to do this myself.
>
> --
> Steve White : Programmer
> Max-Planck-Institut für Gravitationsphysik Albert-Einstein-Institut
> Am Mühlenberg 1, D-14476 Golm, Germany +49-331-567-7625
>
> _______________________________________________
> Developers mailing list
> Developers at cactuscode.org
> http://www.cactuscode.org/mailman/listinfo/developers
>
More information about the Developers
mailing list