[NTLUG:Discuss] sharing hfs+ filesystem on linux

Chris Cox cjcox at acm.org
Wed Jul 20 16:22:24 CDT 2005


MontyS at videopost.com wrote:
> Greetings.
> 
> I have googled all over for this answer, but I can't seem to land on solid
> information.
> 
> I have mounted a HFS+ formatted firewire drive on my fedora core 3 box
> successfully.  I am able to write to the firewire drive, remount it on a
> mac, and everything seems to be working as it should.
> 
> However, I am not able to export the firewire drive via nfs.  Exportfs
> reports the drive as exported, but I get a permissions error when I try to
> connect via nfs from a remote box.  I get the same error from a linux box or
> irix.

Hmmm.. well NFS is very particular... it's not a simple as it seems
at first.

> 
> I wrote an ext3 filesystem on the firewire drive, and was able to export it
> and connect via the remote boxes with no difficulty.
> 
> Looking from another angle, I formatted the firewire drive on the mac as a
> unix drive, whatever flavor of unix filesystem macos uses, but couldn't get
> the firewire drive to mount on my linux box.  >>mount -t ufs ...<< did not
> work.
> 
> I am not able to export firewire drives with fat32 on them either...
> 
> I have not tried samba (smb) as of yet.  My plan is to have a mac formatted
> firewire drive attached to my linux box, which I could export via nfs to
> various irix platforms.
> 
> Any ideas?

Worst case scenario, ditch the kernel based NFS server and go to the
software (user space) one.  It hides a multitude of sins... but you
might be shooting yourself in the foot one day.  The software
based NFS will let you re-export a mounted filesystem for instance..
and other evils.

> 
> TIA,
> 
> Monty
> 
> ps:  Talked to Apple tech support.  They said that their client level os
> wouldn't export any volumes via nfs, and their server level os wouldn't
> export firewire attached volumes.  They couldn't tell me why...
> _______________________________________________
> https://ntlug.org/mailman/listinfo/discuss
> 
> 





More information about the Discuss mailing list