Printing problems and portmapper error related?

[gensoftdes]

Hi, I inherited management of an old SCO Openserver Enterprise system (5.0.5).

The other day it stopped printing...

The jobs were killed and the spooler restarted but no help. Did a shutdown and cold reboot. During the startup process it sat for a long time on "Starting TCP services" but eventually continued. Giving the error "WARNING: portmapper on server (ipaddress) is not responding"

Reading through lots of stuff online I know this error can come up if there is more than one entry for the server on /etc/hosts. I checked and this is not the case.

What's more is that the ipaddress that comes up in the error is completely bogus. Racking my brains I think it might have been the original address of the server a couple years ago, but I'm not sure. What file does the portmapper look at to get the server IP?

Now the printing issue:

The printer is an HP4100N. Setup on the spooler with hpnpcfg. I did try removing the printer and recreating it. The jobs go to the print spooler but they just sit there. They generate a /tmp message file that says:

Current status of printer is:

getone: get_response failed: Connection refused
getone: get_response failed: Connection refused
getone: get_response failed: Connection refused

getone: get_response failed: Connection refused
getone: get_response failed: Connection refused
Unknown status returned by laser.

The printer can be printed to by other pcs on the network. Just not through the unix print spooler.

Help! Any ideas would be most appreciated.

Thank you

 

[motoslide]

I'd swap HPNP with NETCAT. The HPNP sends out an SNMP query before the job is sent. If the printer doesn't answer just right, you will get those "getone" errors.
(Do a google search on "SCO getone" and you will see plenty of helpful docs).
The portmapper issue is normally a problem with multiple entries in /etc/hosts (like you already discovered).
Other places to look include /etc/tcp and /etc/default/tcp.

I think there may be another place if the 5.0.5 has been configured as a web server, but I'm not sure of the details.

It seems reasonable that the portmapper errors could be causing your network print problems as well.

"Proof that there is intelligent life in Oregon. Well, Life anyway.

 

[gensoftdes]

Thanks for the reply. One article I found suggested renaming /etc/getone which I did and it takes care of the error, but the jobs still don't print out.

Also I ran find / -type f -exec fgrep -l '<ip address>' {} \; >/tmp/iplist

for that bogus IP and came up empty.

Can I just chuck the whole darn thing out the window?

 

[PHV]

I'd first check the output of the following command:
ifconfig -a

Hope This Helps, PH.
FAQ219-2884
FAQ181-2886

 

[gensoftdes]

I've checked ifconfig -a and it shows the correct information.

net0: flags=4043<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.2 netmask ffffff00 broadcast 192.168.1.255
perf. params: recv size: 24576; send size: 24576; full-size frames: 1
ether 00:e0:18:54:5a:e0
lo0: flags=4049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 8232
inet 127.0.0.1 netmask ff000000
perf. params: recv size: 57344; send size: 57344; full-size frames: 1

I've also re-run netconfig, rebuilt and relinked the kernel.

I've stopped and started nfs and rpcinit.

If I do rpcinfo -p 192.168.1.2 it gives me all the info.

BTW, I've got the printing working by using netcat instead thanks for the suggestion motoslide.

Now it's just this portmap issue.

Thanks for all the suggestions.

 

[motoslide]

Other than the delay and error on boot, is the system functioning OK?
Ideas:

Do you have consistent information for the system ID?

# uname -X
# hostname
Do these compare favorably with what is listed in your /etc/hosts file?

Do you have an /etc/resolv.conf file? If so, try renaming it, reboot, and see if you still get the portmapper errors.

What are the perms for the /tmp directory?

Do you have just a single NIC in this system?

"Proof that there is intelligent life in Oregon. Well, Life anyway.

 

[gensoftdes]

uname -X and hostname match /etc/hosts

drwxrwxrwt 3 sys sys 6144 Jan 29 17:39 tmp

Yes there is a single NIC in the system.

I'm going to have to try the resolv.conf thing tomorrow.