Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 26 Jan 2007 20:29:28 +0100 (CET)
From:      Oliver Fromme <olli@lurza.secnetix.de>
To:        freebsd-geom@FreeBSD.ORG, rick-freebsd@kiwi-computer.com, fernan.aguero@gmail.com, arne_woerner@yahoo.com
Subject:   Re: clear metadata using dd?
Message-ID:  <200701261929.l0QJTSYR095500@lurza.secnetix.de>
In-Reply-To: <915325.15953.qm@web30303.mail.mud.yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
R. B. Riddick wrote:
 > Rick C. Petty wrote:
 > > Obviously you've never had a disk go bad.  FreeBSD doesn't handle hardware
 > 
 > Obviously u never had a disk go bad, when put under gmirror.

I think he did.  Because he's right:  If a disk dies, the
ata driver detaches it automatically, and gmirror removes
the lost component and continues to run happily in degraded
mode.  No crash.  Of course you should have the disks on
separate cables.  A dying disk sometimes does nasty things
and wedges the whole channel, so you should put the disks
on separate channels.  You can even put them on separate
controllers if you're really paranoid.

Without gmirror, if a disk dies, there's a good chance that
the system will panic or freeze.  Especially when you have
swap on that disk, like the OP seems to have.

 > [...]
 > Does anyone here have SATA disks? I have heard it is easy to pull
 > their plug and plug it back in... Any real test results here?

Here's a real test with a real disk throwing a real error.
:-)

Jan 18 08:03:37 pluto kernel: ad1: TIMEOUT - WRITE_DMA48 retrying (1 retry left) LBA=312581807
Jan 18 08:03:52 pluto kernel: ad1: FAILURE - device detached
Jan 18 08:03:52 pluto kernel: subdisk1: detached
Jan 18 08:03:52 pluto kernel: ad1: detached
Jan 18 08:03:52 pluto kernel: GEOM_MIRROR: Cannot write metadata on ad1 (device=gm0, error=6).
Jan 18 08:03:52 pluto kernel: GEOM_MIRROR: Cannot update metadata on disk ad1 (error=6).
Jan 18 08:03:52 pluto kernel: GEOM_MIRROR: Cannot update metadata on disk ad1 (error=6).
Jan 18 08:03:52 pluto kernel: GEOM_MIRROR: Device gm0: provider ad1 disconnected.

That's from a machine with two SATA drives (ad0 and ad1).
After the incident, the system continued running without
any problems.  gm0 was in degraded mode, using ad0 only.
When the problem was resolved, ad1 was inserted into the
mirror again, resynced, and the box is still running
without any downtime:

Jan 19 12:51:41 pluto kernel: GEOM_MIRROR: Device gm0: provider ad1 detected.
Jan 19 12:51:41 pluto kernel: GEOM_MIRROR: Device gm0: rebuilding provider ad1.
Jan 19 13:45:32 pluto kernel: GEOM_MIRROR: Device gm0: rebuilding provider ad1 finished.
Jan 19 13:45:32 pluto kernel: GEOM_MIRROR: Device gm0: provider ad1 activated.

That machine is running RELENG_6, by the way.

Best regards
   Oliver

-- 
Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M.
Handelsregister: Registergericht Muenchen, HRA 74606, USt-Id: DE204219783
Any opinions expressed in this message are personal to the author and may
not necessarily reflect the opinions of secnetix GmbH & Co KG in any way.
FreeBSD-Dienstleistungen, -Produkte und mehr:  http://www.secnetix.de/bsd

"I learned Java 3 years before Python.  It was my language of
choice.  It took me two weekends with Python before I was more
productive with it than with Java." -- Anthony Roberts



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200701261929.l0QJTSYR095500>