I have been exploring Coda for the past few weeks, and today I am
stumped on a client problem.  When I start venus from any client, it
appears to start up correctly.  But:

---------------------->
11:52:22 /usr/coda/LOG setup for size 0x88008
11:52:22 /usr/coda/DATA initialized at size 0x220020
11:52:23 brain-wiping recoverable store
11:52:24 loading recoverable store
11:52:24 starting VSGDB scan
11:52:24        0 vsg entries in table
11:52:24        0 vsg entries on free-list
11:52:24 starting VDB scan
11:52:24        1 vol entries in table (0 MLEs)
11:52:24        0 vol entries on free-list (0 MLEs)
11:52:25 starting FSDB scan (833, 20000) (25, 75, 4)
11:52:25        0 cache files in table (0 blocks)
11:52:25        833 cache files on free-list
11:52:25 starting HDB scan
11:52:25        0 hdb entries in table
11:52:25        0 hdb entries on free-list
11:52:26 Kernel version ioctl failed (Inappropriate ioctl for device)!
11:52:26 Initial LRDB allocation
11:52:26 Getting Root Volume information...
11:52:26 Venus starting...
---------------------->

It hangs right here, sleeps for about 15 to 20 seconds, then dies.
The venus.log file looks like this:

----------------------->
[ X(00) : 0000 : 11:52:22 ] Coda Venus, version 5.3.6
[ X(00) : 0000 : 11:52:22 ] Logfile initialized with LogLevel = 0 at Wed
Apr 19
11:52:22 2000

[ X(00) : 0000 : 11:52:22 ] E StatsInit()
[ X(00) : 0000 : 11:52:22 ] L StatsInit()
[ X(00) : 0000 : 11:52:22 ] RecovDataSizes: Log = 0x88008, Data =
0x220020
[ X(00) : 0000 : 11:52:23 ] Recov_AllocateVM: allocated 1000 bytes at
20000000
[ X(00) : 0000 : 11:52:23 ] Recov_AllocateVM: allocated 21b000 bytes at
21000000
[ X(00) : 0000 : 11:52:23 ] BeginRvmTruncate (1, 4312, I)
[ X(00) : 0000 : 11:52:24 ] cache::EndRvmTruncate
[ X(00) : 0000 : 11:52:24 ] Recov_DeallocateVM: deallocated 1000 bytes
at 200000
00
[ X(00) : 0000 : 11:52:24 ] Recov_DeallocateVM: deallocated 21b000 bytes
at 2100
0000
[ X(00) : 0000 : 11:52:24 ] BeginRvmFlush (1, 1332, F)
[ X(00) : 0000 : 11:52:24 ] EndRvmFlush
[ X(00) : 0000 : 11:52:24 ] BeginRvmTruncate (2, 1492, F)
[ X(00) : 0000 : 11:52:24 ] cache::EndRvmTruncate
[ X(00) : 0000 : 11:52:24 ] BeginRvmFlush (1, 5516, F)
[ X(00) : 0000 : 11:52:24 ] EndRvmFlush
[ X(00) : 0000 : 11:52:24 ] BeginRvmTruncate (3, 5676, F)
[ X(00) : 0000 : 11:52:24 ] cache::EndRvmTruncate
[ X(00) : 0000 : 11:52:25 ] BeginRvmFlush (1, 430448, F)
[ X(00) : 0000 : 11:52:25 ] EndRvmFlush
[ X(00) : 0000 : 11:52:25 ] BeginRvmTruncate (837, 430608, F)
[ X(00) : 0000 : 11:52:25 ] cache::EndRvmTruncate
[ X(00) : 0000 : 11:52:25 ] BeginRvmFlush (1, 16808, F)
[ X(00) : 0000 : 11:52:25 ] EndRvmFlush
[ X(00) : 0000 : 11:52:25 ] BeginRvmTruncate (1, 16968, F)
[ X(00) : 0000 : 11:52:26 ] cache::EndRvmTruncate
[ X(00) : 0000 : 11:52:26 ] E adviceserver::adviceserver: AdviceServer

[ A(18) : 0000 : 11:52:26 ] adviceserver::main()

[ H(07) : 0000 : 11:52:26 ] HDBDaemon about to sleep on hdbdaemon_sync

[ T(01) : 0000 : 11:52:31 ] BeginRvmFlush (1, 2352, T)
[ T(01) : 0000 : 11:52:31 ] EndRvmFlush

[ D(21) : 0000 : 11:52:31 ] WAITING(SRVRQ):

[ W(20) : 0000 : 11:52:46 ] NotifyUserOfServerDownEvent
[ W(20) : 0000 : 11:52:46 ] DiscoRef= 1

[ D(21) : 0000 : 11:52:46 ] WAIT OVER, elapsed = 14949.2
----------------------------->

I also tried going back to 5.3.5-1, both on the client and on the
server.
The client has the same problem;  venus.log is:

----------------------------->

[ X(00) : 0000 : 08:51:00 ] Coda Venus, version 5.3.5
[ X(00) : 0000 : 08:51:00 ] Logfile initialized with LogLevel = 0 at Wed
Apr 19
08:51:00 2000

[ X(00) : 0000 : 08:51:00 ] ProfInit(): profiling is broken; fix it!
[ X(00) : 0000 : 08:51:00 ] E StatsInit()
[ X(00) : 0000 : 08:51:00 ] L StatsInit()
[ X(00) : 0000 : 08:51:00 ] BeginRvmFlush (1, 60, F)
[ X(00) : 0000 : 08:51:00 ] EndRvmFlush
[ X(00) : 0000 : 08:51:00 ] BeginRvmTruncate (1, 220, F)
[ X(00) : 0000 : 08:51:00 ] cache::EndRvmTruncate
[ X(00) : 0000 : 08:51:00 ] fsobj::Recover: invalid fso (,
(0x7f000001.0x1.0x1))
, attempting to GC...0x2121374c : fid = ((0x7f000001.0x1.0x1)), comp = ,
vol = 2
11b450c
        state = Runt, stat = { 0, 0, -1, -1, 0177777, 255, ??? }, rc
rights = 0
        VV = {[ 0 0 0 0 0 0 0 0 ] [ 0xffffffff 0 ] [ 0 ]}
        voltype = [0 0 1 0], ucb = 1, fake = 0, fetching = 0 local = 0
        rep = 0, data = 0, owrite = 0, era = 1, dirty = 0, shadow = 0
        mvstat = Root
        parent = ((0x0.0x0.0x0), 0), children = 0
        priority = -1 (10967), hoard = [0, -2, 0], lastref = 0
        DisconnectionStatistics:  Used = 0 - Unused = 0 - SinceLastUse =
0
        mle_bindings = (0, 0), cleanstat = [-1, -1]
        cachefile = [ V1, 31677, 0/0 ]
        refs = [0 0 0], openers = [0 0 0]       lastresolved = 0
        discread = 0
[ X(00) : 0000 : 08:51:00 ] BeginRvmFlush (1, 560, F)
[ X(00) : 0000 : 08:51:00 ] EndRvmFlush
[ X(00) : 0000 : 08:51:00 ] BeginRvmTruncate (4, 720, F)
[ X(00) : 0000 : 08:51:00 ] cache::EndRvmTruncate
[ X(00) : 0000 : 08:51:00 ] E adviceserver::adviceserver: AdviceServer

[ A(18) : 0000 : 08:51:00 ] adviceserver::main()

[ H(07) : 0000 : 08:51:00 ] HDBDaemon about to sleep on hdbdaemon_sync

[ D(21) : 0000 : 08:51:05 ] WAITING(SRVRQ):

[ W(20) : 0000 : 08:51:15 ] NotifyUserOfServerDownEvent
[ W(20) : 0000 : 08:51:15 ] DiscoRef= 1

[ D(21) : 0000 : 08:51:15 ] WAIT OVER, elapsed = 10038.7
------------------------------>

Both versions of the clients have been tested (on the corresponding
version of the server) on both RH 6.1 and RH 6.2 machines.  In all
cases, a client that is installed on the same machine as the server
works fine.
**The biggest difference between what I have done today and what I have
done in the past is that the server (either version has the problem) is
running on a machine with RH 6.2, rather than 6.1 (suggesting a 6.2
problem?).  I have not tried reverting to 6.1 and trying again; I wanted
to see if there is a known solution for my problem first.

Thanks,
Gus Scheidt