Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 29 May 1998 12:25:51 +0200
From:      Eivind Eklund <eivind@yes.no>
To:        tcobb <tcobb@staff.circle.net>, "'freebsd-current@freebsd.org'" <freebsd-current@FreeBSD.ORG>, "'freebsd-scsi@freebsd.org'" <freebsd-scsi@FreeBSD.ORG>
Subject:   Re: DPT driver fails and panics with Degraded Array
Message-ID:  <19980529122551.16212@follo.net>
In-Reply-To: <509A2986E5C5D111B7DD0060082F32A402FAD1@freya.circle.net>; from tcobb on Fri, May 29, 1998 at 05:49:30AM -0400
References:  <509A2986E5C5D111B7DD0060082F32A402FAD1@freya.circle.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, May 29, 1998 at 05:49:30AM -0400, tcobb wrote:
> > (1) I've had my array (a 2GB RAID1 - personal RAID :-) run in degraded
> >     mode.  This has worked just fine with the driver in -current, with
> >     the RAID full (of partitions, not data.  I can't understand that
> >     the amount of data should make a difference - the controller
> >     shouldn't know about this anyway).
> 
> Perhaps the difference is RAID-1 versus RAID-5.

This might be so, or there might (I'm probably blaspheming by saying
this) be a difference or bug in the handling of degradation from DPT
to DPT.

What is obvious is only that you're having problems with a particular
controller/driver/kernel combination, and that a part of the driver has
problems coping with failure somewhere else.  I don't think it would
be wise to consider the problem to be narrowed down more than that.

Now, to be able to debug this as effectively as possible I suggest
you/we try to create a list of test-cases that we believe would lock
down the problem:

Hypothesis 1: The problem occur when using your type of controller,
   RAID5+HotSpare, "large" amounts of capacity used, your kernel, and
   a disk fail.
Verification: Create above setup, try to fail a disk on it.

Now, which parameters to vary depend on whether we can get the above
to crash.  It might be dependent on having exactly equal disks, too
:-(

Eivind.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980529122551.16212>