Re: [RFC PATCH] mountd: export: Deal with NFS filesystems

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Jan 31, 2022 at 06:15:41PM +0100, Richard Weinberger wrote:
> ----- Ursprüngliche Mail -----
> > Von: "bfields" <bfields@xxxxxxxxxxxx>
> >> Using /proc/fs/nfsfs/volumes it is possible to find the NFS fsid
> >> from the backend and use it as seed for mountd's UUID mechanism.
> > 
> > Sorry, I haven't checked--what is that number, really?  It's probably
> > not the fsid returned from the server, as that wouldn't be guaranteed
> > unique across multiple servers.  But that means it's probably generated
> > in a way that doesn't guarantee it's stable across reboots.  And we need
> > filehandles to work across reboots.
> 
> Unless I badly misunderstood the code it comes from fs/nfs/client.c's
> nfs_create_server() where the NFS client fetches NFS_ATTR_FATTR via getattr().
> So it should be unique:
> https://datatracker.ietf.org/doc/html/rfc7530#section-5.8.1.9

Unique within one server, but not across different servers.

Depending on how your servers assign fsids, collisions might be quite
likely.  Results of a collision would be pretty bad.

--b.



[Index of Archives]     [Linux Filesystem Development]     [Linux USB Development]     [Linux Media Development]     [Video for Linux]     [Linux NILFS]     [Linux Audio Users]     [Yosemite Info]     [Linux SCSI]

  Powered by Linux