[Simh] KS10 and the DMR
simh at swabhawat.com
simh at swabhawat.com
Thu May 1 17:30:42 EDT 2014
Phil,
Thank you for your observations and remarks.
To give you a better impression with more details, see the following:
OpenVMS Network status for local node 63.2 SWBA01 on 1-MAY-2014 20:30:08.33
Node Links Cost Hops Next Hop to Node
63.2 SWBA01 0 0 0 (Local) -> 63.2 SWBA01
63.31 SWBU01 0 4 1 EWA-3 -> 63.31 SWBU01
63.33 SWBU03 0 4 1 EWA-3 -> 63.33 SWBU03
63.41 SWBX01 0 4 1 EWA-3 -> 63.41 SWBX01
63.51 SWBW01 0 4 1 EWA-3 -> 63.51 SWBW01
63.52 SWBW02 0 4 1 EWA-3 -> 63.52 SWBW02
63.54 SWBW04 0 9 2 EWA-3 -> 63.89 SWBV89
63.55 SWBW05 0 9 2 EWA-3 -> 63.89 SWBV89
63.56 SWBW06 0 9 2 EWA-3 -> 63.89 SWBV89
63.57 SWBW07 0 9 2 EWA-3 -> 63.89 SWBV89
63.59 SWBW09 0 9 2 EWA-3 -> 63.89 SWBV89
63.89 SWBV89 0 4 1 EWA-3 -> 63.89 SWBV89
63.168 SWBP68 0 4 1 EWA-3 -> 63.168 SWBP68
Total of 13 nodes.
[ANF10 network: connected to ANFW04(53), located at ANFW04(53), 6 nodes]
Node ANFW04 (53) 56(8), 55(8), 54(8)
Node ANFW05 (54) 60(8), 57(8), 53(8)
Node ANFW06 (55) 53(8)
Node SWBW07 (56) 53(10)
Node SWBW08 (57) 54(10)
Node ANFW09 (60) 54(8)
[DECnet network: local node SWBW04, 13 reachable nodes in area 63]
Name Number Line Cost Hops L.Links Delay
SWBA01 (63.2) DMR-0-0 6 2
SWBP68 (63.168)DMR-0-0 6 2 1 1000
SWBU01 (63.31) DMR-0-0 6 2
SWBU03 (63.33) DMR-0-0 6 2
SWBV89 (63.89) DMR-0-0 2 1
SWBW01 (63.51) DMR-0-0 6 2
SWBW02 (63.52) DMR-0-0 6 2
SWBW04 (63.54) local 0 0 0 5000
SWBW05 (63.55) DMR-1-0 2 1
SWBW06 (63.56) DMR-2-0 2 1
SWBW07 (63.57) DMR-0-0 7 2
SWBW09 (63.59) DMR-3-0 2 1
SWBX01 (63.41) DMR-0-0 6 2
You can see tiny differences here; so by the way the circuit costs are
differring between Vms and Tops10; in Vms the Dmr costs 5, the Una/Ewa costs
4.
On the Tops10, the Dmr costs only 2 while the Una/Ewa costs are 4 (6-2 of
the Dmr). The Una is of course faster than the Dmr by far.
The machines here on this Decnet list vary from Tops10, Tops20, Vms,
Rsx11MP46, RSTS 10-1L to Win98.
Dec got a lot of things right; in my times they held a dominant position at
the universities here, especially in the laboratories and workgroups and
they held it by rights.
Concerning software versions this was a first try of what to choose.
As it was a niche product with almost none to scarcely a few in the field
and the number of years now passed, it was a safe guess that V017 was never
released but was only inhouse with Tim on the road to V022. V022 was
probably the last because of the errors contained in it; had these been
corrected in a later version for a client in the field, it would certainly
have appeared on the TSU tapes. It did not. So V017 and V023 are initially a
safe guess with little chance of collision at present even if another
inhouse V017 may have existed.
I do not know what the official Dec version policy was at the time (if any)
but it can easily be changed as it is not yet spread around.
Then D8RINT.MAC is not the only one changed; the same proble4m is there as
for 7.04 upwards monitor module SYSINI.MAC had to be adapted as well and
there are 2 versions of that one: for 7.04 release and for 7.05 through TSU
updates.
So I am open for suggestions ..
Another thing to be ironed out is concurrent Dmr and Kdp; momentarily it
will not run because of OVA crash; OVA is Overflow Virtual Addresses.
The KS10 is very memory constrained and a lowly configured system for 63
processes, 16 terminal and 16 remote lines and 8 Dmr's (the extra lines are
cheap, a few pages extra) has only about 430 pages of core left.
Stop codes were usually formed from the letter abbreviations of the stop
text (PDLOVL=Push Down List OVerfLow).
Now what should WEM stand for? You may be very well right there as these
kinds of pranks were practiced at Dec in those days.
It is by the way a most non-descriptive code as it is used as the garbage
bin for all the kinds of fatal network errors that could not be classified
otherwise; but with a deadly ring to it. Death is there but so to speak you
have to start a search party first for locating the corpse to start a
post-mortem ..
This probably is the reason for the VMS copy error:
%COPY-E-OPENIN, error opening SWBW04"xxxxxx
xxxxxx"::DSKB:[10.7.MON]D8RV23.LST as input
-RMS-E-ACC, ACP file access failed
-SYSTEM-F-REMRSRC, insufficient system resources at remote node
I have seen it working only once by chance; all the KS10's have this same
problem and wildcarding the same thing.
On the other hand, the same operation to the larger KLH10's with >6000 pages
but the same versions of FAL have none of these problems.
With NFT there are no such problems, not even with wildcarding. It looks
like VMS needs a lot of memory on the FAL or otherwise.
I have yet to find a way to debug the FAL process for that DAP error as it
is started up in the Galaxy/Opr system; a version started up from the prompt
does not work as the incoming network requests are not fielded to it. I
suppose DDT should work in some way on the detached port FAL runs on.
All in all a nice trip down memory lane as once I also managed one of the
first combinations of a Tops20 Decsystem-2060 with a Vax8650 on a Delni ,
with an ethernet Decserver 100 and a printer on a Decserver 150; and then
via the yellow Ethernet cable segment to some PC's with Depca Ethernet card
with Decnet; all phase-IV already.
It was also the time of witty users from the States bypassing their allotted
limited number of async terminal mux ports to the Dec20, by abusing the Vax
Vms to do set hosting to the Dec20 to get there; over a satellite line then.
That route was quickly cut off though ..
It was the time of the clearinghouse tapes from Grenoble. Quite adventurous
and humorous when looking back.
Best regards.
Reindert
-----Original Message-----
From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-edge.com]
On Behalf Of Phil Budne
Sent: Thursday, May 01, 2014 17:25
To: simh at trailing-edge.com
Subject: Re: [Simh] KS10 and the DMR
> From: "R. Voorhorst" <simh at swabhawat.com>
> Date: Thu, 1 May 2014 08:10:23 +0000 (UTC)
.....
> The Ks10 Dmr case hos now been solved; the Dmr can be used in Tops10
> V7.02 to 7.05 with Decnet and Anf, but some minor issues have still to
> be resolved.
> The trace can be followed at the Simh issue list #100:
> https://github.com/simh/simh/issues/100
Thanks for the detailed notes!
It's wild to see the virtual network you created; I was just thinking this
week about _some_ things that DECnet got right (backwards compatible
evolution thru "Phases", PHase IV "areas" could span over multiple "links";
"hello timer" values in routing packets)
.network /topology
[ANF10 network: connected to BITXT1(10), located at BITXT1(10), 1 node]
Node BITXT1 (10) None
[DECnet network: local node BITXT1, 12 reachable nodes in area 7]
Name Number Line Cost Hops L.Links Delay
BITXOM (7.72) KDP-0-0 7 2
BITXOO (7.61) KDP-0-0 4 1
BITXOR (7.71) KDP-0-0 7 2
BITXOT (7.70) KDP-0-0 7 2 0 1000
BITXOU (7.82) KDP-0-0 7 2 0 1000
BITXOV (7.60) KDP-0-0 7 2 0 1000
BITXOW (7.74) KDP-0-0 7 2 0 1000
BITXOX (7.76) KDP-0-0 7 2
BITXOZ (7.81) KDP-0-0 7 2
BITXT0 (7.79) KDP-0-0 7 2
BITXT1 (7.80) local 0 0 0 5000
MBSERV (7.6) KDP-0-0 7 2
But....
That delivers the following:
1. Dmr Tops10-702: D8RINT V006 ANF ok Decnet phase-III ok
2. Dmr Tops10-703: D8RINT V016 ANF ok Decnet phase IV
crash
3. Dmr Tops10-704/5: D8RINT V022 ANF ok Decnet phase IV
crash (UBA)
......
The things to do are:
1. Retrieve either BB-X128A-SB or a full backup of a 702 system pack
with [10,7] build tree
2. Repair D8RINT.MAC V016 to a newer V017 and possibly some Tops10-703
modules
3. Repair D8RINT.MAC V022 to a newer V023 and possibly some Tops10-704
modules
V017 of the file was already used (on the way to V020 (if versions were
octal, or V018 if they weren't)), and there's some possibility that there
was a V023 that wasn't released, or hasn't been found, yet.
So V016A and V022A might be better designations for modified files.
Also I have some tiny memory that "WEM" stopcode was one where it was
someone's initials! Wayne Matson?
http://pdp-10.trailing-edge.com/bb-bt99g-bb/01/d60unv.mac.html
; [111] 12-May-81 WEM Add D6IAB, D6OAB error returns
Phil
_______________________________________________
Simh mailing list
Simh at trailing-edge.com
http://mailman.trailing-edge.com/mailman/listinfo/simh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20140501/17cebb50/attachment-0002.html>
More information about the Simh
mailing list