[OpenAFS] strange things happen when adding an additional database server (4 of 4)

[Paul Blackburn]

Hello,

I have just installed OpenAFS 1.2.8 on a RedHat 8.0 machine (single-homed).

When adding an additional AFS database server something odd happens.
I restarted {ka,bu,pt,vl}server on every database server in the cell.

DNS resolution seems OK. Loopback interface is up.
All 4 database servers are correctly time-synch'ed.

The new (4th) database server has not joined in with the UBIK vote and
has not synchonized  with copies of the database files.

Any ideas?

# cat /usr/afs/logs/AuthLog
/usr/afs/bin/kaserver: problems with host name Ubik init failed
l database.
Mon Mar 10 20:15:52 2003 Using level crypt for Ubik connections.
Mon Mar 10 20:15:52 2003 ubik: primary address 127.0.0.1 does not exist

# host 127.0.0.1
1.0.0.127.in-addr.arpa domain name pointer localhost.

# tail -4 /usr/afs/logs/BosLog
Mon Mar 10 20:20:21 2003: kaserver exited with code 2
Mon Mar 10 20:20:21 2003: BNODE 'kaserver' repeatedly failed to start, 
perhaps missing executable.
Mon Mar 10 20:20:21 2003: kaserver exited with code 2
Mon Mar 10 20:20:21 2003: BNODE 'kaserver' repeatedly failed to start, 
perhaps missing executable.

# bos status localhost -i kaserver -long
Instance kaserver, (type is simple) temporarily disabled, stopped for 
too many errors, currently shutdown.
    Process last started at Mon Mar 10 20:20:21 2003 (13 proc starts)
    Last exit at Mon Mar 10 20:20:21 2003
    Last error exit at Mon Mar 10 20:20:21 2003, by exiting with code 2
    Command 1 is '/usr/afs/bin/kaserver'

# ls -l /usr/afs/bin/kaserver
-rwxr-xr-x    1 root     root       270424 Dec 11 21:07 
/usr/afs/bin/kaserver

--
cheers
paul                                http://acm.org/~mpb