Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 21 Apr 1998 08:42:02 -0500
From:      Karl Denninger  <karl@mcs.net>
To:        Stephen McKay <syssgm@dtir.qld.gov.au>
Cc:        freebsd-current@FreeBSD.ORG
Subject:   Re: NFS corruption
Message-ID:  <19980421084202.01633@mcs.net>
In-Reply-To: <199804210559.PAA09123@troll.dtir.qld.gov.au>; from Stephen McKay on Tue, Apr 21, 1998 at 03:59:47PM %2B1000
References:  <199804210559.PAA09123@troll.dtir.qld.gov.au>

next in thread | previous in thread | raw e-mail | index | archive | help
Mount with the "-2" switch (NFSv2) and see if the problem disappears.

We saw this one, and John Dyson is aware of it.  It has NOT been fixed,
however, at least as of this time.

It is access-pattern dependant - 99% of the patterns you can come up with
will not break.  However, "make world" will, and so will trying to do things
like modifying the mail system to write user mailboxes to their home
directores.

This was a BITCH to track down far enough to guess that mounting with NFSv2
would have an effect.

--
-- 
Karl Denninger (karl@MCS.Net)| MCSNet - Serving Chicagoland and Wisconsin
http://www.mcs.net/          | T1's from $600 monthly / All Lines K56Flex/DOV
			     | NEW! Corporate ISDN Prices dropped by up to 50%!
Voice: [+1 312 803-MCS1 x219]| EXCLUSIVE NEW FEATURE ON ALL PERSONAL ACCOUNTS
Fax:   [+1 312 803-4929]     | *SPAMBLOCK* Technology now included at no cost

On Tue, Apr 21, 1998 at 03:59:47PM +1000, Stephen McKay wrote:
> NFS is eating my .depend files during a make world.
> 
> The client is a Compaq Prolinea 486SX33 with 12Mb ram and 2 small SCSI-1
> disks on an Adaptec 1542b.  The server is a pentium 133 with 32Mb ram and
> 2 4Gb IBM UW SCSI disks on a FirePort 40 (ncr 875).
> 
> The source and objects are on the server and NFS mounted to the client.  The
> client mounts src readonly and obj read-write (and async, if that does
> anything under NFS).
> 
> The client kernel (and userland) is -current from April 19 (April 18 US time),
> and has DIAGNOSTIC set.  The server is -current from March 12, just before
> the big VM changes.
> 
> Memory is short on the client, so paging is brisk.  There is plenty of
> swap space free.  I don't run CAM or softupdates.
> 
> I ran 'make -j2 buildworld' and several hours later observed unusual error
> messages complaining about garbage in .depend files.  Many .depend files
> were affected.  Each .depend file was broken similarly.  They would start
> normally, then the corruption would start on a page boundary (multiple of
> 0x1000), but *not* extend as far as the next page boundary.  The corruption
> was either C source, or C preprocessor output overwriting the normal contents.
> 
> >From this evidence I suspect bugs when handling partial page writes, either
> in NFS or the general VM system.  I can't be sure whether the client or the
> server is at fault.
> 
> Unfortunately, this is a pretty vague diagnosis.  Fortunately, this is now
> a permanent setup at my place, so I can run tests until my power bill
> bankrupts me!
> 
> So, any ideas on how I should tackle this?
> 
> Stephen.
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-current" in the body of the message

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



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980421084202.01633>