From: Andreas Gruenbacher Subject: Re: Native NFSv4 ACLs on Linux Date: Fri, 14 Jul 2006 17:16:59 +0200 Message-ID: <200607141716.59827.agruen@suse.de> References: <200607141637.32246.agruen@suse.de> <1152888586.14027.18.camel@lade.trondhjem.org> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-15" Cc: nfs@lists.sourceforge.net, nfsv4@ietf.org Return-path: To: Trond Myklebust In-Reply-To: <1152888586.14027.18.camel@lade.trondhjem.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: nfsv4-bounces@ietf.org List-ID: On Friday, 14 July 2006 16:49, Trond Myklebust wrote: > On Fri, 2006-07-14 at 16:37 +0200, Andreas Gruenbacher wrote: > > Hello, > > > > This is to announce a prototype that implements NFSv4 ACLs natively on > > Linux. So far, the implementation supports NFSv4 ACLs on Ext3 > > filesystems. The code is functional, but hasn't seen a whole lot of > > testing so far. > > Ooh. Interesting... > > Are you planning on encoding the acl model that is being used directly > in the on-disk filesystem flags instead of relying on a mount option? I haven't thought about this much so far. At the moment it's either POSIX ACLs or NFSv4 ACLs, per filesystem, and even if xattrs of the "other" ACL model exist on the filesystem, they will be completely invisible. We could encode the ACL model used in the filesystem flags, or think about mixed models as known from NetApp Filers. There is a lot more work to be done before we'll get there though. > The incompatibilities between the two models is bound to cause trouble > if some administrator messes up the mount command line. Quite likely, yes. Andreas _______________________________________________ nfsv4 mailing list nfsv4@ietf.org https://www1.ietf.org/mailman/listinfo/nfsv4