Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 11 Sep 2024 13:02:06 +0000
From:      bugzilla-noreply@freebsd.org
To:        fs@FreeBSD.org
Subject:   [Bug 281417] Inconsistent restrictions on jailed and sharenfs properties affecting NFS in a VNET jail
Message-ID:  <bug-281417-3630-1UuCSytXz4@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-281417-3630@https.bugs.freebsd.org/bugzilla/>
References:  <bug-281417-3630@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D281417

--- Comment #4 from Oliver Kiddle <okiddle@yahoo.co.uk> ---
(In reply to Rick Macklem from comment #3)

The 'zoned' property is directly analagous to the FreeBSD 'jailed' property.
And in this case, it is the 'jailed' property that I have set. I then use
zfs-jail(8)/unjail from the vnet jail's exec.created/exec.prestop so the
datasets are only mounted and controlled from inside the jail.

Setting sharenfs is rather easier and more flexible than /etc/exports. As I
mention it is possible to set jailed on a parent dataset and sharenfs on a
child. The error message and attempts to treat the two properties as mutual=
ly
exclusive are thus incompletely applied. And while that could be treated as=
 the
bug making the two properties work together would be more useful. I've just
realised that I did have a manual entry in /etc/exports so was not really u=
sing
the property.

With jailed NFS, it should perhaps instead be /etc/zfs/exports within the j=
ail
that is modified for a jailed dataset.

I've been trying different combinations in an effort to get rpc.rquotad to =
work
from the jail - something I've not as yet been successful with. It needed z=
fs
allow -e just for zfs userspace to work in the jail.

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-281417-3630-1UuCSytXz4>