Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

SCO 5.0.5 network services stop

0 views
Skip to first unread message

David Baker

unread,
Mar 28, 2002, 2:40:01 PM3/28/02
to
I have a client with a Compaq Proliant 1600 with SCO OS Enterprise 5.0.5.
The network services stop around the same time each day. The serial
services are still running and I can login via modem on a multiport serial
board or com port.

I have tried ...
SCO software patches per SCO Technical Articles
Pacific Code Works Tune Up software
Manually looked at some different parameters in the kernel with no luck
Replaced the NIC with a 3COM 3c905 PCI and new drivers
Replaced the hub that the server connects to
Erased the system configuration with Compaq Smart Start and reinstalled SCO
from scratch

The problem persists. Does anyone have any suggestions?

--


David Baker
Account Manager
Net-Tech
1121 Trotwood Ave. Suite 10
Columbia, TN 38401
(931) 540-1188 phone
(931) 540-1199 fax
www.getnetworked.com

"Business Computer Networking Solutions"

Bill Vermillion

unread,
Mar 28, 2002, 3:06:04 PM3/28/02
to
In article <SnKo8.17036$1f1.1...@e3500-atl2.usenetserver.com>,
David Baker <dba...@getnetworked.com> wrote:

>I have a client with a Compaq Proliant 1600 with SCO OS Enterprise
>5.0.5. The network services stop around the same time each day. The
>serial services are still running and I can login via modem on a
>multiport serial board or com port.

>I have tried ...
>SCO software patches per SCO Technical Articles
>Pacific Code Works Tune Up software
>Manually looked at some different parameters in the kernel with no luck
>Replaced the NIC with a 3COM 3c905 PCI and new drivers
>Replaced the hub that the server connects to
>Erased the system configuration with Compaq Smart Start and reinstalled SCO
>from scratch

Have a question for you? Or make that several questions.

By 'stop' do you mean the tcp/ip services are not running
as in the S85tcp stop. Does S85tcp start get it running?

It sounds like you are grabbing at straws trying to find the
problem instead of the bottom up network trouble shooting approach.

Does the local interface work?
If no you have a SW problem.
If yes does the local ip work
If no you have a configuration problem
Does a remote ip work
If no but the other pass then you probably don't have a problem in
that machine and need to look elsewhere.

In all trouble shooting start at the lowest level and work upward
and NEVER skip a step.
--
Bill Vermillion - bv @ wjv . com

David D. Chuha

unread,
Mar 28, 2002, 9:49:23 PM3/28/02
to
David Baker wrote:
> I have a client with a Compaq Proliant 1600 with SCO OS Enterprise 5.0.5.
> The network services stop around the same time each day. The serial
> services are still running and I can login via modem on a multiport serial
> board or com port.
>
> I have tried ...
> SCO software patches per SCO Technical Articles
> Pacific Code Works Tune Up software
> Manually looked at some different parameters in the kernel with no luck
> Replaced the NIC with a 3COM 3c905 PCI and new drivers
> Replaced the hub that the server connects to
> Erased the system configuration with Compaq Smart Start and reinstalled SCO
> from scratch
>
> The problem persists. Does anyone have any suggestions?

I had a similar problem with a Unixware 7 machine. Networking services
would just quit for a while. The time that this would happen wasn't
predictable but I could bet that it would happen at least once a day. I
tried many things to fix the problem.

The last two things I tried were upgrade the drivers for the NIC and
kill routed. Since you've already upgraded your NIC drivers I would try
killing routed (unless this machine is actually routing anything). I
would also try stopping any other unnecessary services.

--
Dave

Jeff Liebermann

unread,
Mar 31, 2002, 12:57:39 PM3/31/02
to
On Thu, 28 Mar 2002 13:40:01 -0600, "David Baker"
<dba...@getnetworked.com> wrote:

>I have a client with a Compaq Proliant 1600 with SCO OS Enterprise 5.0.5.
>The network services stop around the same time each day.

I recently walked a potential client (while I was still in the
hospital) over approximately the same situation and symptoms. Every
day, at approximately noon, one server would magically be inaccessible
from his workstations. The workstations could see and use other
servers, but not this one particular server. Workstations already
connected to this server would remain connected, but no new
connections could be made.

What happened was the CFO would fire up his misconfigured W2K laptop
at about noon and grab his daily dose of spam. Some IT expert had
cleverly assigned two IP addresses to his network card. One was the
correct address, but the 2nd one was the IP address of the server.
OSR5 was duitfully complaining about duplicate IP's but nobody noticed
as the server console was inaccessible and there were no complaints
logged in syslog or /usr/adm/messages. A simple:
ipconfig /all
on the W2K laptop showed the problem.

Since your problem seems to happen at the same time of day, it's easy
enough to fire up a network sniffer on a different machine and plug it
into the same port as the OSR5 server. (Use a hub if you're plugged
into a switched port). I'm partial to Ethereal on Linux(RH7.x) this
week. Netmon for NT4 will also work. Log only the traffic to or from
the server and ignore everything else. You should see traffic
instantly stop at some point. Scroll back a few packets and you'll
probably see what was causing the problem and possibly whom to blame.


--
Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
(831)421-6491 pgr (831)426-1240 fax (831)336-2558 home
http://www.LearnByDestroying.com WB6SSY
je...@comix.santa-cruz.ca.us je...@cruzio.com

Jeff Liebermann

unread,
Apr 1, 2002, 11:25:15 PM4/1/02
to
On Sun, 31 Mar 2002 09:57:39 -0800, Jeff Liebermann
<je...@comix.santa-cruz.ca.us> wrote:

>On Thu, 28 Mar 2002 13:40:01 -0600, "David Baker"
><dba...@getnetworked.com> wrote:
>
>>I have a client with a Compaq Proliant 1600 with SCO OS Enterprise 5.0.5.
>>The network services stop around the same time each day.

(...)

One more possibility. If the routing table gets trashed by
bogus RIP update, it could similate something that looks like
network services stopping (whatever that means). Run:
netstat -rn
and see if it makes sense. If it's trash, and you are NOT using
RIP, kill the "routed" daemon and reboot. Edit the file:
/etc/tcp
and comment out the section that starts routed.

0 new messages