cause of NFS error (EBADFD(81)) and others

[Gafling]

I am looking for a guru who can delineate the reasonable causes of the huge number of '/var/adm/syslog' messages:

Jan  2 09:47:52 xxxxxxxx WARNING: NLM: 4 request to server 110.22.30.8 failed: 8 (errno 118)

These messages are usually followed by a message like:

Jan  2 09:49:14 xxxxxxxx last message repeated 3345 times

On this same SCO system, 'ls -l /remotedir' will sometimes return

/remotedir file handle is in a bad state(81)

BTW, if one were to log in on the remote system, '/remotedir' would be found and permissions are 777. So the object of the 'ls -l' exists and is accessible.

The inconsistant behaviour is frustrating efforts to isolate the problem to either my SCO OS506 system or the remote system which is an NCR 4400 running MP/RAS Unix 3.02. The list of suspects are of course the network, the NICs, etc. I tend to not think these are to blame because these are 100Mbit segments; plenty of bandwidth. Other suspects are the lock daemons and other inetd related software on either system.

Scanning the SCO website returns NO information for 'EBADFD' and very little information regarding 'nfs' at all. SO any and all input is welcome.

gafling

 

[PHV]

Explanation for the errno 118 :
[ENOPROTOOPT] (118) Protocol not available
A bad option or level was specified when getting or setting options for a protocol.
HTH
PH.

 

[Gafling]

'PHV': thank you for the response. You must work for MicroSoft Tech Support as your answer is totally correct but provides no real help. Seriously, I need some tech insight on why the daemons on either system do not seem to get along at times. Is there a version issue such as NFS version x.x is supported on one system but NFS version y.y is supported on the other. Are the lock daemons compatible? Has SCO made changes to NFS recently?

I have proposed this issue to SCO tech support but I suspect my children will be grown before I get an answer.

gafling