Subject: Re: dump(8) behaviour
To: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
From: Manuel Bouyer <firstname.lastname@example.org>
Date: 10/07/2004 20:47:08
On Thu, Oct 07, 2004 at 08:27:39PM +0200, Hauke Fath wrote:
> >Could it be an issue with metadata being changed on the filesystem under
> >dump's feets ? With amanda dumps may run a bit slower (because he runs
> >several in parallel), and so such a race condition is more likely to happen.
> Not too likely. As for Tom, my Amanda's dump seems to hang on the biggest
> (server) partitions. And that is a data partition (NetBSD source and other
> stuff) that is idle at night.
> I do not recall it hanging on /var.
I've never seen it hang, but for now I only have one 2.0 system backed up with
amanda (it's a sparc SMP system, with only a 4GB disk). All others are 1.6.x.
However I see quite often dump failing on a live /var/mail partition.
Manuel Bouyer <email@example.com>
NetBSD: 26 ans d'experience feront toujours la difference