Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 02 Aug 2001 19:48:48 +0200
From:      "Julian Stacey" <jhs@jhs.muc.de>
To:        freebsd-hackers@freebsd.org
Cc:        David Malone <dwmalone@maths.tcd.ie>, Ian Dowse <iedowse@maths.tcd.ie>
Subject:   Re: vnconfig + mount removes permission for a second 
Message-ID:  <200108021748.f72HmmE05676@jhs.muc.de>
In-Reply-To: Your message of "Thu, 02 Aug 2001 11:48:35 BST." <20010802114835.A92803@walton.maths.tcd.ie> 

next in thread | previous in thread | raw e-mail | index | archive | help
David Malone wrote:
> On Wed, Aug 01, 2001 at 11:35:32AM +0000, Julian Stacey wrote:
> > So something is momentarily making the image unreadable.
> > Should FreeBSD [mount/kernel ?] be changed to avoid denying access ?
> 
> When you do a mount it automatically HUP's mountd which then
> re-exports NFS filesystems. I suspect what is happening is that
> the the filesystem mountlist is being cleared for a moment and that
> is upsetting the cp.
> 
> You could try moving /var/run/mountd.pid out of the way and running
> the mount command. That should stop mount HUPing mountd, which
> would let you test this theory.

Thanks,
I tried your suggestion, it works ! so Mountd is the problem then !

Cp does not fail on Umount, I guess Umount happens quicker.
  ( One reason might have been that Umount is silent, whereas Mount caused 7
    lines of complaint to console EG
	mountd[96]: bad exports list line /cdrom -alldirs -ro -mapall
    (20 years ago Unix console errs came out under polling not interrupt,
     so that might have been a lengthy time window, but I expect things
     on FreeBSD & xconsole are now improved :-)
    But I commented out /cdrom in /etc/exports, & still see Mount breaking Cp.
  )

I can also reproduce this behaviour on just one host as both amd/nfs
host & target (thus avoiding ethernet, except possibly for named via ethernet)


> Ian Dowse <iedowse@maths.tcd.ie> wrote:
> 
> See PRs misc/3980 and kern/9619 for more details. I think NetBSD
> tried at one point to make mountd incrementally change the export
> list, but it turned out to be quite hard to get the logic right to
> keep the mountd and kernel lists in sync. I think they reverted
> that change eventually.

I haven't seen PRs misc/3980 and kern/9619, is there any info is this thread
that should be added to one of those ?

> This is certainly a bug that needs to be fixed; mountd should be
> able to build up a list of all exports for a filesystem and pass
> them into the kernel in one "replace export list"  operation.
> 
> Maybe nice'ing mountd to run at a higher priority,

Question is how high ? With Nice or Rtprio or something else ?

>  and/or specifying
> only IP addresses in /etc/exports would help things a bit now.

I converted my /etc/exports to entirely numeric,
threw a sighup at named & mountd, but that did not stop the problem.
Killed mountd -r & restarted with numeric exports, still did not help.

Julian
J.Stacey  Munich Unix (FreeBSD, Linux etc) Consultant  http://bim.bsn.com/~jhs/
   Ihr Rauchen => mein allergischer Kopfschmerz !  Schnupftabak probieren !

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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200108021748.f72HmmE05676>