[OpenAFS] How to access files on the server without the client?

[Stephen Joyce]

  This message is in MIME format.  The first part should be readable text,
  while the remaining parts are likely unreadable without MIME-aware tools.

--43577093-219805381-1210163702=:14596
Content-Type: TEXT/PLAIN; CHARSET=X-UNKNOWN; FORMAT=flowed
Content-Transfer-Encoding: QUOTED-PRINTABLE
Content-ID: <Pine.LNX.4.64.0805070835181.14596@hellbender.physics.unc.edu>

I know it's not what you want to hear, but as far as I know, you can only=
=20
access that data via the fileserver.

Whenever I've been in a similar situation in the past, it always helps to=
=20
try to figure out what changed since the last time the server booted=20
cleanly.

You said you had a power outage that caused this problem; did all the=20
filesystems, particularly the one with the openafs binaries on it, come up=
=20
cleanly? If you didn't have any file corruption, the problem likely isn't=
=20
caused by the power outage, but by something else. [I've had problems show=
=20
up on a reboot after a power outage not because of the power outage, but=20
because servers get rebooted seldomly while in production.]

Can you can do a md5 checksum of all the openafs-related files=20
(binaries and conf files) on the problem server and another server of the=
=20
same architecture and running the same version/build of openafs to locate=
=20
what's changed? You might want to re-run through your install docs and=20
either double-check everything, or else re-install and re-configure=20
everything, copying known-good files over the ones on your server...? Just=
=20
make sure that you've got a copy of all of your relevant config files=20
backed up, or that you can get them from another server (CellServDB,=20
KeyFile, UserList).

Before any of that, if you're running iptables or some other firewall on=20
the server or the test client, turn it off for testing to see if it's a=20
problem...

[Of course, as a last-ditch you can reinstall the problem fileserver,=20
taking care not to wipe your /vicep partitions. Or since you said you=20
had 20 servers, simply move the /vicep disks from the problem server to=20
another fileserver. Procedures for normalizing vldb after the latter type=
=20
of move have been discussed on this list before.]

Best of luck!

Cheers, Stephen
--
Stephen Joyce
Systems Administrator                                            P A N I C
Physics & Astronomy Department                         Physics & Astronomy
University of North Carolina at Chapel Hill         Network Infrastructure
voice: (919) 962-7214                                        and Computing
fax: (919) 962-0480                               http://www.panic.unc.edu

In pioneer days they used oxen for heavy pulling, and when one ox couldn't=
=20
budge a log, they didn't try to grow a larger ox. We shouldn't be trying=20
for bigger computers, but for more systems of computers.
                             --Rear Admiral Grace Hopper

What would you rather have to plow a field - two strong oxen or 1,024=20
chickens?                   --Seymour Cray



On Wed, 7 May 2008, Micha=B3 Dro=BCdziewicz wrote:

> Hi,
> Is it possible to access volumes (files on volumes) without the client? I=
 can=20
> log in onto the server as root.
>
> Server is not serving volumes (problem I've described erlier - fileserver=
 and=20
> volserver are not starting properly) and I really need to copy these file=
s=20
> onto another machine for this server to reinstall. Any help would be grea=
tly=20
> appreciated.
>
> --=20
> xmpp/email: koniczynek@uaznia.net
> xmpp/email: koniczynek@gmail.com
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
>
--43577093-219805381-1210163702=:14596--