Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 17 Apr 2003 13:12:39 -0600 (MDT)
From:      Fred Clift <fclift@verio.net>
To:        Ian Dowse <iedowse@maths.tcd.ie>
Cc:        freebsd-stable@freebsd.org
Subject:   Re: PATCH: Forcible delaying of UFS (soft)updates 
Message-ID:  <20030417130651.N46464-100000@vespa.dmz.orem.verio.net>
In-Reply-To: <200304162310.aa96829@salmon.maths.tcd.ie>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 16 Apr 2003, Ian Dowse wrote:

>
> Just after the disk spins up there is a small window where the
> cached writes get written out in a burst. Due to the amount of
> cached data and the probable re-ordering of writes, the disk is
> quite likely to be in an inconsistent state during this flurry of
> writes, but the window is short so it is probably not a big issue
> in practice.

Of course, this is when your power-supply is most likley to fail due to
the sudden increased load :).

<Anecdote>
I lost a disk that I had been occasionally using as a backup drive due to
an effect like this.  I had two scsi drives in an external enclosure, and
I wanted to re-newfs the other drive in the enclosure so I started a tar
job to copy all the files over and the PS blew about 20 seconds into the
write since both drives were 'busy' rather than just one or the other as
had been the case for quite a while as the machine sat in the corner and
did nothing for a year.  The target drive was hosed bad enough that you
couldn't newfs it any more and the vendor's low-level format tools claimed
the disk was unrepairable...
</Anecdote>

I guess in a laptop, this failure mode isn't as likley as in my case...


Fred

--
Fred Clift - fclift@verio.net -- Remember: If brute
force doesn't work, you're just not using enough.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030417130651.N46464-100000>