[OpenAFS] Re: afs vs nfs

[Neulinger, Nathan]

Oh... Very cool. Did not realize that at all...

-- Nathan

------------------------------------------------------------
Nathan Neulinger                       EMail:  nneul@umr.edu
University of Missouri - Rolla         Phone: (573) 341-6679
UMR Information Technology             Fax: (573) 341-4216
=20

> -----Original Message-----
> From: Derek Atkins [mailto:warlord@MIT.EDU]=20
> Sent: Wednesday, November 23, 2005 10:22 AM
> To: Neulinger, Nathan
> Cc: openafs-info@openafs.org
> Subject: Re: [OpenAFS] Re: afs vs nfs
>=20
> "Neulinger, Nathan" <nneul@umr.edu> writes:
>=20
> > Something as simple as having an alternative namei that=20
> would look the
> > same down below, but would have the top level directory be=20
> the RW volume
> > id corresponding to whatever data is being stored.
>=20
> Um, the top-level directory (well, two top-level dirs) already is
> (are) the volume id.  It's just encoded in base64, and split into
> two sets of bits.  But it's still the volumeID.  E.g.
>=20
> ./+/++++U/+/+/0++++6
> ./1/1+++U/special/zzzzPA++++0
>=20
> These are in two different volumes.
>=20
> > This would allow the top level dirs and volume header to be=20
> backed up
> > and restored as regular files. Granted you would not have the true
> > snapshot like you can get with vos, but even that is=20
> subject to timing
> > as far as high level file access goes.=20
>=20
> Current namei already gives you that, actually.
>=20
> -derek
> --=20
>        Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
>        Member, MIT Student Information Processing Board  (SIPB)
>        URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
>        warlord@MIT.EDU                        PGP key available
>=20
>=20