Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 15 Feb 1999 10:10:32 +0000 (GMT)
From:      Doug Rabson <dfr@nlsystems.com>
To:        Dag-Erling Smorgrav <des@flood.ping.uio.no>
Cc:        Dag-Erling Smorgrav <des@FreeBSD.ORG>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG
Subject:   Re: cvs commit: src Makefile.inc1 src/release Makefile          src/sys/alpha/conf Makefile.alpha src/sys/i386/conf Makefile.i386          src/sys/pc98/conf Makefile.pc98 src/usr.bin/chpass Makefile          src/usr.bin/passwd Makefile
Message-ID:  <Pine.BSF.4.05.9902151009050.50431-100000@herring.nlsystems.com>
In-Reply-To: <xzpemntt6iy.fsf@flood.ping.uio.no>

next in thread | previous in thread | raw e-mail | index | archive | help
On 14 Feb 1999, Dag-Erling Smorgrav wrote:

> Dag-Erling Smorgrav <des@FreeBSD.ORG> writes:
> >   Log:
> >   Ignore errors from chflags. This makes it possible to make installworld
> >   with DESTDIR set to an NFS-mounted file system.
> 
> OBTW, I tested this by installing the bin distro from 3.0-RELEASE on a
> scratch filesystem, then NFS-mounting it over lo0 and running make
> installworld with the NFS mount point as DESTDIR. When using NFSv3,
> the nfsd process would eventually hang in the nfsrcv state, and the
> rest of the system would exhibit the classical "NFS server is dead"
> symptoms. I had to reboot to get unwedged. It finally worked (albeit
> slowly) when I forced an NFSv2 mount.

The nature of NFSv3's two stage write-commit protocol, combined with
interactions with the buffer cache (which is used by both the client and
server end of the transaction) produce an unavoidable deadlock for
loopback mounts.  Don't do this.

--
Doug Rabson				Mail:  dfr@nlsystems.com
Nonlinear Systems Ltd.			Phone: +44 181 442 9037



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



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.05.9902151009050.50431-100000>