Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 8 Jan 2000 21:22:30 -0500 (EST)
From:      "Crist J. Clark" <cjc@cc942873-a.ewndsr1.nj.home.com>
To:        mb@imp.ch (Martin Blapp)
Cc:        freebsd-bugs@FreeBSD.org
Subject:   Re: bin/14709: umountall requests possibly mishandled by mountd(8)
Message-ID:  <200001090222.VAA10117@cc942873-a.ewndsr1.nj.home.com>
In-Reply-To: <Pine.SGI.4.10.10001072147260.6248-100000@mephisto.imp.ch> from Martin Blapp at "Jan 7, 2000 09:52:45 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
Martin Blapp wrote,
> 
> Hi,
> 
>  
> > Not so fast.
> > 
> > Jan  7 09:18:25 backmail mountd[144]: umountall request from
> > 192.168.229.3 from unprivileged port
> > 
> > It still is griping about umountall requests coming from itself.
> > 
> > # uname -a
> > FreeBSD backmail.mydom.org 3.4-STABLE FreeBSD 3.4-STABLE #0: Tue Dec 21
> > 14:23:45 EST 1999
> > postman@backmail.mydom.org:/u1/FreeBSD-3S/src/sys/compile/BACKMAIL  i386
> 
> the umntall request is sent by another box ! So I bet you've still got
> another 3.3 BSD box and you use 192.168.229.3 as NFS-Server for it. I
> fixed the broken client behaviour and nothing else.
> 
> hope this explains it.

But looking back at the original PR, this is the exact same
problem. My NFS server, backmail (192.168.229.3) generated the
following messages yesterday,

Jan  7 08:22:00 backmail mountd[144]: umountall request from 192.168.229.3 from unprivileged port
Jan  7 09:18:25 backmail mountd[144]: umountall request from 192.168.229.3 from unprivileged port

I can correlate all of these messages to messages on another
machine. This machine gets syslog messages from timed when a machine
on the network disappears because it has been shutdown (trimmed so
that there is only one message per machine per shutdown, timed will
produce three before giving up),

Jan  7 08:24:55 newmail timed[94]: measure: lucy.mydom.org did not respond
Jan  7 09:19:57 newmail timed[94]: measure: curly.mydom.org did not respond

It just so happens that each of these machines is an IRIX (5.3, 6.5,
and 6.2, respectively) and none of them ever actually mount
filesystems from the FreeBSD server, backmail.

All of this is contained in the original PR and the mail which it
references. From my point of view (I am no NFS expert), the problem
must be fixed with mountd and not clients. Fixing the FreeBSD client
does not help me because the machines causing the messages are not
FreeBSD machines. If the IRIX clients are doing something odd that
should be logged, then I think the fix needs to be to make that error
message useful. Saying the request came from itself is either
incorrect (it should say it came from the client), or it is at best,
useless (why bother with that portion of the message if it always is
reporting itself as the origin of the request) or at worst,
misleading. 
-- 
Crist J. Clark                           cjclark@home.com


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200001090222.VAA10117>