Date: Fri, 23 Apr 93 04:30:08 PDT
From: Advanced Amateur Radio Networking Group <tcp-group@ucsd.edu>
Errors-To: TCP-Group-Errors@UCSD.Edu
Reply-To: TCP-Group@UCSD.Edu
Precedence: Bulk
Subject: TCP-Group Digest V93 #106
To: tcp-group-digest
TCP-Group Digest Fri, 23 Apr 93 Volume 93 : Issue 106
Today's Topics:
ARP- A Routing Problem?
NOS and PC's turbo switch..... (3 msgs)
X.500 service
Send Replies or notes for publication to: <TCP-Group@UCSD.Edu>.
Subscription requests to <TCP-Group-REQUEST@UCSD.Edu>.
Problems you can't solve otherwise to brian@ucsd.edu.
Archives of past issues of the TCP-Group Digest are available
(by FTP only) from UCSD.Edu in directory "mailarchives".
We trust that readers are intelligent enough to realize that all text
herein consists of personal comments and does not represent the official
policies or positions of any party. Your mileage may vary. So there.
----------------------------------------------------------------------
Date: Thu, 22 Apr 93 12:22:57 +0100
From: A.D.S.Benham@bnr.co.uk
Subject: ARP- A Routing Problem?
To: TCP-Group@UCSD.Edu
I've found something strange with ARP over the past few days.
I run JNOS108d with a 2m and a 70cm port - the 2 ports have different IP
addresses (the 70cms port is not available very often, but when it is it
provides better links. Therefore neighbouring stations can use routing metrics
to try the 70cms port first, and if it isn't available use 2m instead).
Everything was fine until one of my neighbouring hubs had equipment failure,
and changed to a backup system that has only a 70cms port.
He now puts out ARP requests for 'g8fsl-2m' on 70cms. I receive these
requests on 70cms, and send him an ARP reply... BUT with the source ip
address set to g8fsl-70cm. He doesn't see this as a reply to his request,
and even if he had ARP eavesdrop on it wouldn't give him the ARP information
that he needs to resolve his request for 'g8fsl-2m'.
Looking at the source code, if the ARP request is for an IP address used by
one of the interfaces at the station, then the 'request source' data is copied
to the 'reply target' data, and the 'reply source' data is obtained from the
iface structure for the interface on which the ARP request was received. I had
always naively assumed that the source and target fields were just exchanged.
What I don't know is whether it's a bug or a feature. Anybody any ideas?
The problem will appear to go away when my neighbour restores his real
equipment, but will it pop up again in the future?
73,
Andrew Benham
--------------------------------------------------------------------
adsb@bnr.co.uk BNR Europe Ltd, London Road, Harlow, Essex CM17 9NA
adsb@bnr.ca +44 279 402372 Fax: +44 279 402100
Home: g8fsl@g8fsl.ampr.org [44.131.19.195] G8FSL@GB3XP
--------------------------------------------------------------------
------------------------------
Date: 22 Apr 93 07:45:58 CDT
From: Jack Snodgrass <kf5mg@vnet.IBM.COM>
Subject: NOS and PC's turbo switch.....
To: <TCP-Group@UCSD.Edu>
I've found that when I exit NOS, my box has slowed down. When I run
norton's SI before I start, I have an SI rating of 100. I then start
NOS and exit immediately. SI now shows a 50 rating. Has anyone observed
this before? I've played with not using my ethernet board, and not using
ansi.sys instead of nansi.sys, but it doesn't seem to make a difference.
If I run NOS without my autoexec.net file, the speed doesn't change. I
haven't gone through and commented stuff out yet. I was hoping someone
had seen this before. Thanks.
73's de Jack - kf5mg
AMPRnet - kf5mg@kf5mg.ampr.org - 44.28.0.14
AX25net - kf5mg@kf5mg.#dfw.tx.usa.na - work (817) 962-4409
Internet - kf5mg@vnet.ibm.com - home (817) 488-4386
------------------------------
Date: Thu, 22 Apr 93 13:28:00 CDT
From: Ben Thornton <ben@teton.sps.mot.com>
Subject: NOS and PC's turbo switch.....
To: kf5mg@vnet.IBM.COM
>
> I've found that when I exit NOS, my box has slowed down. When I run
> norton's SI before I start, I have an SI rating of 100. I then start
> NOS and exit immediately. SI now shows a 50 rating. Has anyone observed
> this before? I've played with not using my ethernet board, and not using
> ansi.sys instead of nansi.sys, but it doesn't seem to make a difference.
> If I run NOS without my autoexec.net file, the speed doesn't change. I
> haven't gone through and commented stuff out yet. I was hoping someone
> had seen this before. Thanks.
>
One (remote) possibility is that you have noise on one of your
serial port wires and this could be generating a bunch of spurious
interrupts. I shouldn't do this since NOS is supposed to be
disabling the UART's interrupt line when it exit()'s, so it
is not very likely. However when writing comm code in the past
I have encountered noise on dangling modem handshake wires which
can cause many interrupts to occur...
--ben
--
Ben Thornton Amateur callsign: WD5HLS
Internet: ben@teton.sps.mot.com Motorola, Inc.
Amprnet: wd5hls@wd5hls.ampr.org Austin, TX
<<These are my opinions and not those of my employer.>>
------------------------------
Date: Thu, 22 Apr 93 14:10:58 -0700
From: karn@qualcomm.com (Phil Karn)
Subject: NOS and PC's turbo switch.....
To: Jack Snodgrass <kf5mg@vnet.IBM.COM>, <TCP-Group@UCSD.Edu>
At 07:45 AM 4/22/93 CDT, Jack Snodgrass wrote:
> I've found that when I exit NOS, my box has slowed down. When I run
>norton's SI before I start, I have an SI rating of 100. I then start
>NOS and exit immediately. SI now shows a 50 rating.
This is bizarre. I don't reprogram the hardware timer (the likely cause
of this symptom) in my code, but it's possible that somebody else's
version does. Whose code are you running?
Phil
PS. Greetings from Dayton.
------------------------------
Date: Thu, 22 Apr 1993 15:50:44 -0500 (CDT)
From: Steve Sampson <ssampson@sabea-oc.af.mil>
Subject: X.500 service
To: TCP-Group@UCSD.Edu
ron@edu.wayne.eng.chaos writes:
> Besides, I very rarely even waste my time with AX.25 packet anymore since I
> do TCP/IP, and some kind of server that can handle internet style Email
> addresses and AX.25 addresses, that is accessible in real-time via internet
> gateways, and non-real time by regular packet, seems like a better idea.
I suspect you mean that you use TCP/IP *over* AX.25, since the current rules
don't allow transport of third-party data using anything else. The current
BBS addressing scheme doesn't really do much, other than direct the mail to
a particular state and hostname. The other parts of the address can be thrown
out as they serve no purpose.
I've been thinking, wouldn't it be easier to convert the BBS programs to use a
domain name server or database of that style. eg:
domain.txt:
w9u.ga. IN CNAME w9u.bbs.ampr.org.
w9u.bbs.ampr.org. IN A 44.99.99.99
(or a fictional BBS rather than IN system):
w9u.ga. BBS CNAME w9u.bbs.ampr.org.
w9u.bbs.ampr.org. BBS MX 10 wb5fff
w9u.bbs.ampr.org. BBS MX 20 wd8ggg
Using 'rewrite' for all the necessary conversions. And then routing it via an
IP route table or the given BBS address in domain.bbs above.
That will give them time to make the conversion, and then you can delete the
BBS style address 5 years down the road, when they are each given an IP address
to automate the bit masking and routing.
There's probably not a lot of BBS systems in each state, so a database of host
names can't be very large. Just remove the state name and put bbs.ampr.org in
its place. The number of HF forwarding stations is even smaller, so those
would probably be the "best neighbors" in any routing tables.
Normally an user or BBS shouldn't be concerned with what everyones full
address is, only which neighbor knows the BBS being addressed. If someone
puts an address in as "jim@w9u.bbs.ampr.org" or any variation of w9u, then
the domain database gets the routing address. That would mean the HF BBS's
would need to exchange updates as new BBS systems come up or pass away.
> (I'm about 2-3 hops from the WP server AD8I and it takes anywhere
> from a couple hours to a couple days for a response. Totally unacceptable.)
Your BBS should just route the data to the nearest known neighbors in each
direction, and punt the other unknown addresses to the forwarding node (HF or
Satellite). I don't think we need to DNS lookup the address before sending it
on a low speed network system (<= 1200 bps). If we can't reach the address
directly then the forwarding node certainly can, or it will return it.
This doesn't sound very dynamic, but BBS's don't really change that much. Most
of the NOS systems should rely on a forwarding node to do it's thing, and not
become a BBS themselves. They could use one NOS system to import and export
BBS mail, but the other hosts should then use their local SMTP or POP mailers
to this system. Is this too KISS'y?
---
Steve, N5OWK
------------------------------
End of TCP-Group Digest V93 #106
******************************
******************************