Hi!

The AFS manual says "It is a good practice to place a read-only volume at
the read/write site, for a couple of reasons." and a little later "The
other reason to place a read-only volume at the read/write site is that
the Cache Manager does not attempt to access the read/write version of a
replicated volume if all read-only copies become inaccessible. If the file
server machine housing the read/write volume is the only accessible
machine, the Cache Manager can access the data only if there is a
read-only copy at the read/write site."

I encountered the following today:

Case #1)

File and DB server A holds both RW and RO volumes root.afs, root.cell and X
File and DB server B holds RO copies of root.afs, root.cell and X
File and DB server C holds no relevant volumes
Client D has just written a file Y in volume X

Now, A drops out of the network and all data in the cell becomes
inaccessible. For ever. We waited over 30 minutes for any possible
timeouts to fire, but no. Every operation just returns a "connection timed
out".

Case #2)

File and DB server A holds the RW volumes root.afs, root.cell and X
File and DB server B holds RO copies of root.afs, root.cell and X
File and DB server C holds no relevant volumes
Client D has just written a file Y in volume X

Now, A drops out of the network; for a while all data on X is inaccessible
(as is to be expected) and after some time, D can access the RO volume
from B. B can also access the data on its own RO volume, but C cannot! I
waited for over 15 minutes, but it still could not.

What is wrong here? The network outage was, as you probably guess, caused
precisely in order to test this kind of failsafe operation and consisted
of a disconnected RJ45 plug on A and nothing else.

I was under the impression that this kind of failsafe operation in case
one fileserver is lost, is the heart of AFS. I also thought we had
everything set up exactly as the manual recommends.

Cheers,
Juha

-- 
                 -----------------------------------------------
                | Juha Jäykkä, [EMAIL PROTECTED]                        |
                | Laboratory of Theoretical Physics             |
                | Department of Physics, University of Turku    |
                | home: http://www.utu.fi/~juolja/              |
                 -----------------------------------------------

Attachment: pgp7HfwnTQMZk.pgp
Description: PGP signature

Reply via email to