Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!samsung!zaphod.mps.ohio-state.edu!van-bc!rsoft!mindlink!a1040
From: a1040@mindlink.UUCP (Robert Broughton)
Newsgroups: comp.dcom.lans
Subject: Help! (Novell Netware/286)
Message-ID: <2803@mindlink.UUCP>
Date: 11 Aug 90 00:49:13 GMT
Organization: MIND LINK! - British Columbia, Canada
Lines: 25

I'm working with a Novell Netware/286 network using Puredata ARCNet
cards (16 bit on the server, 8 bit on the workstations). It was
installed two months ago. It worked fine for the first month, but has
been plagued with glitches since; Corrupted records, corrupted FoxBase
indexes, "Internal Consistency Error" reported by FoxBase when
re-indexing or sorting.

There is one specific symptom that I want to ask about; If you run
FCONSOLE, and display the statistics summary, the value shown for
maximum "routing buffers" is 100 (We've also used values of 40 and 150).
In the second column, peak routing buffers used this session, the figure
will be 65,535(!). The third column is also bizarre. When the system is
brought up, the routing buffers currently in use will be slightly under
65,535, and will decrease at the rate of about 5 per second if there is
light activity on the system, or about 200 per second if there is heavy
activity on the system. When this figure gets down to zero, it will wrap
around to 65,535, and continue to drop. This will continue until the
server is brought down.

Has anyone out there seen this phenomenon? I've talked to several local
experts, and one in Provo, Utah, and none of them have seen it. One
other thing: This is version 2.15 C of Netware/286.

Robert Broughton  -  a1040@mindlink.uucp
                -or- a1040%mindlink@wimsey.bc.ca