[prev in list] [next in list] [prev in thread] [next in thread] 

List:       freebsd-qa
Subject:    Re: NFS problem?
From:       "Brian J. McGovern" <bmcgover () cisco ! com>
Date:       2003-03-12 12:42:31
[Download RAW message or body]

I have had one solid lockup in the i386<->i386 case during heavy NFS writes
while moving large files (ISO images of CDs) back to the server after upgrading
to 5.0 (client was also 5.0).

I didn't have a debugger in the kernel at the time, and haven't reproduced it
since I put it in to get a back trace/dump (~2 weeks now). I was waiting to 
post on the topic until I had some data to provide useful debugging, and to
make sure it wasn't a fluke (new hardware, as well).

I'll let the lists know if I see this again.

 > Hi,
 > 
 >  I experienced an NFS problem between an i386 box and a sparc64 box.
 >  The i386 box that is running 5-CURRENT as of Mar 7 acts as an NFS server,
 >  and the sparc64 box that is running 5-CURRENT as of Mar 5 acts
 >  as an NFS client.  The problem is that the server is locked up
 >  under heavy loads such as doing make release.  Once locked,
 >  it is not responsive to outcoming ssh connection request and so on,
 >  and I had to reboot the box...
 > 
 >  A friend of mine told me in an i386(server)<->i386(client) case
 >  it was OK under heavy loads, and I confirmed that in an
 >  i386(server, 4-STABLE)<->sparc64(client, 5-CURRENT) case it also worked fin
e.
 >  As far as I can check, the problem seems to happen in an
 >  i386(server, 5-CURRENT)<->sparc64(client, 5-CURRENT) case only.
 > 
 >  I don't know what is the trigger, but did anyone experience
 >  the same problem? 
 > 
 > -- 
 > | Hiroki SATO  <hrs@eos.ocn.ne.jp> / <hrs@FreeBSD.org>
 > 
 > To Unsubscribe: send mail to majordomo@FreeBSD.org
 > with "unsubscribe freebsd-sparc" in the body of the message

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-qa" in the body of the message
[prev in list] [next in list] [prev in thread] [next in thread]