[OpenAFS] 'replication'

[Turbo Fredriksson]

I've been writing a script that looks for all volumes that
should be replicated (http://www.bayour.com/scripts/update_afs.sh)
but I'm _NOT_ happy.

First of, I'm running both the server and client on only one machine,
so maybe this is the problem.

The thing is that when it's replicating a 'big' volume (1190595 K)
I get timeouts on my OWN volume (and all the other users also get
timeouts to their volumes). And this is NOT the volume that's being
replicated. I can 'understand' that when I'm replicating my volume,
I 'loose' contact with it while it's being cloned (I'm not very
happy with this either - it happens when I'm doing a dump as well).

The other thing is that it takes forever to replicate the 7.6Gb in 
the 8 volumes I'm replicating, so there is no way I can run it every
10 minutes...

And doing it once every night, well then the replica sites are outdated
most of the time!


And I also wonder why it want's to do a complete release of my own
volume (it's mounted, does that matter?) but not on the other (not
mounted).


UPDATE: I've been running this almost all day (as soon as the script
ended, I started it again). Running in verbose mode, no replica volume
mounted, I see that ALL volumes is a complete release of the volume'.

Can't I do a incremental release. I'm _NOT_ specifying '-force', so
I can't see why it's doing a complete release... Is the 'vos help
release' wrong? 

What did I do wrong?

-- 
fissionable explosion Khaddafi Cuba bomb KGB president nitrate quiche
Treasury security plutonium [Hello to all my fans in domestic
surveillance] 767 ammonium
[See http://www.aclu.org/echelonwatch/index.html for more about this]