[OpenAFS] Problem with starting back up

[Volker Lendecke]

--G4iJoqBmSsgzjUCe
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Tue, Mar 08, 2005 at 09:00:04AM -0700, Dexter Kimball wrote:
> Take a look at ports in use -- we've seen this on a Solaris box running s=
ome
> sort of RAID management utility and IBM/Transarc AFS -- it was using
> 2040/tcp -- turns out the volserver talks to the fileserver on this port =
=2E..
> FSYNC_clientInit comes from fileserver/volserver talking to each other ab=
out
> volume headers IIRC.
>=20
> (If I have the port number wrong, anyone, please correct.)

I've come across 2040/tcp just yesterday. Would it be feasible to use a unix
domain socket on the systems that provide them? This would avoid any port
collisions and could be nicely protected to be accessed only by root.

Looking over fssync.c a patch looks not too difficult.

Volker

--G4iJoqBmSsgzjUCe
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFCLdLnP/hMgvhmbo4RAo6SAJwKdjdcYJV8FXnmUGiXGgxntYDQfQCfdFxD
XtUjiGaQBNNFiNF9e5SbPJA=
=5Tu/
-----END PGP SIGNATURE-----

--G4iJoqBmSsgzjUCe--