HLC illegal port useage
Sean White
I'm sitting with a bit of a sticky problem. The default setup for using an
HLC on a 2200-series system is for it to use TCP port 102, which is assigned
by the IANA for ISO-TSAP useage. TCP/IP in the DCP is also using an 'illegal'
port (port 256) but this is not too bad as it hasn't been assigned by the IANA
yet.
I have essentially two questions. First, can an alternative port be specified
for the HLC and secondly, are Unisys planning on legalising their useage of
ports with the IANA?
--
DISCLAIMER: These opinions are *mine* - if you want my boss' opinion,
go and ask him yourself... :-)
James R. Heit
I can somewhat address the "HLC" part of this. I claim to have little or
no knowledge of DCPs (and I'm right!). For starters, HLCs are not
directly tied to TCP. Other protocols can be run over HLCs. When you
say the HLC uses TCP port 102, that is an incorrect statement. The 2200
software that runs the HLC is CMS 1100. If you are using TCP/IP protocol
over your HLC, CMS 1100 will do a TCP passive open on well-known port 102
on behalf of TP0. TP0 is an OSI protocol. OSI applications can be run
over TCP/IP networks using TP0. This is CMS 1100's implementation of RFC
1006. RFC 1006 specifies the use of TCP port 102 for this purpose.
Now that I've said all that, I add the following disclaimers:
-I'm not the expert. If you wish to contact the Unisys experts in this
area, write a Contact.
-I'm not an official spokesman. This message reflects my opinions only.
--
James R. Heit
Principal Software Engineer
2200 Products Group
Host Communications Software
Unisys Corp.
jr...@rsvl.unisys.com
Sean White
: HLC on a 2200-series system is for it to use TCP port 102, which is assigned
: by the IANA for ISO-TSAP useage.
It has been pointed out to me that my understanding of the interaction between
CMS1100 and the HLC is incomplete. It would be more accurate to say that I have
misunderstood the setup completely.
: secondly, are Unisys planning on legalising their useage of
: ports with the IANA?
I would like to withdraw my query - I should have researched the issue better
before making what is tantamount to an accusation in a public forum. My posting
was ill-considered.
I apologise to Unisys, the readers of and contributors to this newsgroup
for my conduct.
Helge Moulding
: >[Something about ports 102 and 256...]
Telcon's TCP-IP Stack uses the following assignments, all legal:
Type Port Number (decimal) Authority
==== ===================== =========
Telnet 23 RFC1700
Domain Name System 53 RFC1700
RFC1006 TP0 102 RFC1700
Telnet "passthrough" 256,257,258 RFC1700 (*)
RIP 520 RFC1700
(*) "unassigned" -- RFC1700 indicates that the range of "assigned"
ports used to be 0-255, which left 256... up for grabs. "Recently,"
it says, the range has been expanded up to 1023.
I don't think this is a problem, since Telnet "passthrough"
is not the preferred method. Instead, use the TCP-IP Stack
implementation of RFC1006 (since TCP-IP Stack 2R2).
: >I have essentially two questions. First, can an alternative port be specified
: >for the HLC and secondly, are Unisys planning on legalising their useage of
: >ports with the IANA?
HLC? Check your CMS configuration manuals. Legalization? Heck, everything
is as legal as church on Sunday! The problem you seem to be struggling
with is what protocols you will use to communicate between 2200 and
DCP. As long as your Telcon release levels and CMS release levels are
compatible, I know of no problems if you just use the default port
assignments.
James R. Heit wrote,
: I can somewhat address the "HLC" part of this. [...]
What he said.
--
Helge Moulding
Helge Moulding
: I apologise to Unisys, the readers of and contributors to this newsgroup
: for my conduct.
Shoot, Sean, if you can't shoot your mouth off on newsnet, what is
it good for? Don't sweat it. This was one question I was actually
able to contribute to, so I didn't mind. But it was a right nice
apology, a good example for any out there who have trespassed.
--
Helge Moulding
Sean White
: Shoot, Sean, if you can't shoot your mouth off on newsnet, what is
: it good for? Don't sweat it.
Thank-you... shooting my mouth off is exactly what I should *not* have done,
though. Like I said, it wasn't very well thought out.
: This was one question I was actually
: able to contribute to, so I didn't mind.
I learned quite a bit from the responses, as well as from reading RFC1006. Now
I'm curious as to the actual protocols used. Where can I find more
documentation on these protocols? Are they publicly available?
Best regards,
Sean.
Helge Moulding
: I'm curious as to the actual protocols used. Where can I find more
: documentation on these protocols? Are they publicly available?
If you want, you can send me private email. For any reading here
to whom the answer is of interest, as well, all RFCs are available
by anonymous ftp at ds.internic.net. I don't remember the directory
off hand, but it's easy enough to find. There are also FAQs available
on a lot of this stuff, at rtfm.mit.edu.
Douglas Comer has authored a three volume book (the last co-authored)
on these protocols, _Networking with TCP-IP_. There are a couple
other good books, as well, but Comer's books are what I am familiar
with.
--
Helge "Hope this helps" Moulding
James R. Heit
>Helge Moulding (h...@unislc.slc.unisys.com) wrote:
>
>: Shoot, Sean, if you can't shoot your mouth off on newsnet, what is
>: it good for? Don't sweat it.
>
>Thank-you... shooting my mouth off is exactly what I should *not* have done,
>though. Like I said, it wasn't very well thought out.
>
>: This was one question I was actually
>: able to contribute to, so I didn't mind.
>
>I learned quite a bit from the responses, as well as from reading RFC1006. Now
>documentation on these protocols? Are they publicly available?
>
> Sean.
>--
>(swh...@gov.za)
>
>DISCLAIMER: These opinions are *mine* - if you want my boss' opinion,
> go and ask him yourself... :-)
I assume you are asking for more information about OSI session,
presentation, and application protocols. Yes, I think they are publicly
available, but I'm not sure how to get a hold of a copy of these docs. I
have a copy of ISO9542, and it was printed in Switzerland. My impression
is that these are not as easily available as Internet RFCs. Also, since
OSI is not as widely used as TCP/IP, I would expect the documents to be
more uncommon. If anyone knows of an Internet source for these, please
post.