🚀 go-pugleaf

Sorry, I tried to send email on this one and it bounced, and I don't want
to wade thru 50 articles to find the original...

In article <hippo.34.0@sonoma.edu> you write:
>We are using a Network General Sniffer with a 3c505 card. The setting are
>IRQ 11, DMA 6 I/O 0300. For some reason when I load the packet driver for
>this card it takes about 45 seconds for it to load. In addition I can not
>get IPX to load the sniffer Frezzzzzzzzzzzzeee. The IPX is from BYU and is
>working with packet driver on all other PC's. The Packet Driver is version
>10. Any ideas ?

I don't use a sniffer, but the lanalyzer will not allow you to run the
software on top of a packet driver.  The Lanz software *must* be the only
thing talking to the board.  Loading ipx after lanz or vice versa will
freeze the machine.  To load ipx after lanz, you have to reboot, then
reboot again to use the other.

I suspect that the sniffer works the same way...
-- 
Shava Nerad Averett				shava_averett@unc.edu
/*  all materials (c)1992, Shava Nerad Averett, and have nothing significant
    to do with the University of North Carolina, a mostly owned subsidiary 
    of the NC Legislature, a mostly owned subsidiary of the DOT.      */

In article <1992Sep3.230016.27803@samba.oit.unc.edu> shava@hermes.oit.unc.edu (Shava Nerad Averett) writes:
>
>I don't use a sniffer,

Good show! Oops, sorry...

>but the lanalyzer will not allow you to run the
>software on top of a packet driver. The Lanz software *must* be the only
>thing talking to the board.

Actually, a good chunk of the LANalyzer code does run down on the board.
Assuming you have an EXOS 325 (the hardware component of the "traditional"
LANalyzer), there is a 286 processor and 2 megabytes of memory down
on the hardware. It runs an executive, if that's the right word, that
handles all the buffer management, and filtering and some of the statistical
analysis functionality. Packet drivers just talk to relatively "dumb"
hardware. When running the executive, the EXOS 325 is a veritable
nuclear physicist compared to most Ethernet boards and the conversations
that occur between the card and the host code occur at a much higher level.

>Loading ipx after lanz or vice versa will freeze the machine.  To load ipx 
>after lanz, you have to reboot, then reboot again to use the other.

Really? I've not had that problem myself although it is not something
I really expect to work so I don't get upset if it goes belly up on me.
Recently I've been getting into the LANalyzer for NetWare which is
a software only Network Analysis tool. It uses ODI drivers and "regular"
hardware to do it's thing. One very remarkable feature it has (that
certainly had my chin hitting the desk when I discovered it) is that
even though it has to slam the LAN Adapter into promiscuous mode to
do it's packet capturing and analysis job, both NetWare and LAN WorkPlace
keep on truckin' just like everythings normal! One minute you are watching
the LANalyzer for NetWare dashboard and all the stats being updated on
the screen and the next minute you have clicked the Windows file
manager up onto the screen and are poking around with drive F: and a minute
later you are typing characters at a UNIX system using the LAN WOrkPlace
"Host Presenter" Windows telnet client and everything just keeps
humming along. I have since evolved a (probably incorrect) theorum about how 
this is pulled off but even so, someone over in our building 6 deserves a 
serious round of applause for some criminally cunning codemanship.

Keith
-
Keith Brown                                           Phone: 408-473-8308
Novell San Jose Development Center                      Fax: 408-473-8990
2180 Fortune Drive, San Jose, CA 95131                  Net: keith@novell.COM

>
>>I am soon to implement a DAT tape backup system for a Novell 3.11 lan of
>>approx. 50 users and am looking for software recommendations/opinions.
>
>DAT will be dead within two to 3 years as a backup medium - Why not look
at 
>an industry-standatd EXABYTE tape drive....
> | / | Jon Biddell              Email: Jon.Biddell@unsw.EDU.AU      | \ |

Why will DAT be dead?  Exabyte is tied to a single manufacturer and
will continue to be very expensive.  DAT is 5-10 gigabytes for $1500. 
File
access in 30 seconds.  Why would it die?

Julian Humphries
Cornell University
  ^                                                                  ^

In article <lacnu8INNoat@phad.hsc.usc.edu> kmartin@phad.hsc.usc.edu (Kenneth J. Martin) writes:

>Subject says it all.  We are getting stuck shift,ctrl, and alt states on
>our IBM clone workstations.  This really came to bang when one of the 
>lab stations had both a stuck ctrl and alt state at the same time, and then
>someone pressed the delete key and guess what, the computer rebooted (gee, I 
>wonder why).

>Here is what we are running:
>Qemm386.sys RAM x=d000-dbff st:m
>        *note: we tried using the IA parm, but it didn't work
>wd8003e (9.7.0)
>byuipx 
>emsnetx (3.26)

>Has anybody else had these problems?  Anybody have a solution? 
>Is it QEMM? Is it the packet driver?  Is it the motherboards from Twinhead?


Are you sure it's version 9.70 of the packet driver.  Version 10 wd8003e.com
has a bug with that identical problem.  The fix has been published.  I don't 
know if Russ Nelson has released version 11 yet with the fixed binaries 
though.  It only takes about 10 minutes to fix the bug if you have the asm 
source code and Borlands Turbo Assembler.

 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 John Tymann                  || INTERNET   : tymann@ops1.bwi.wec.com
 Westinghouse Electric Corp.  || (W) Vaxes  : tron::"tymann@ops1"
 (410) 765-3915               || Compuserve :  72230,3327
  KBones Heal. Pain is Temporary. Chicks Dig Scars. Glory is Forever!@