T.R | Title | User | Personal Name | Date | Lines |
---|
5985.1 | Same here | BIKINI::KRAUSE | CSC Network Management/Hubs | Wed May 11 1994 05:31 | 4 |
| I just tried it here - and have the same problem. Is anybody able to
investigate this (I don't have the time to do it now)?.
*Robert
|
5985.2 | More info
| WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Wed May 11 1994 08:12 | 35 |
| Hi, Robert..
Thanks for the report
I looked at the trace between the MCC station and the OSF machine while I
was trying to register it, and observed the following sequence repeated a
number of times
MCC station makes a connect
It sends a get request
The OSF system responds
MCC station closes the connection.
When the registration problem occurs the MCC station makes a connection,
sends a request and the OSF machine sends a disconnect.
using my (very) limited knowledge of CMIP it looks as though the Module which
MCC was trying to query the OSF machine about at the time was code 24 Hex
which according to the MCC dictionary corresponds to 'DUA'.
This would indeed seem to be the offending item as issuing the command
'show node xxx dua' from the MCC command line results in the error report
Communication with the target has been interrupted
Entity Existence Info = Entity Exists
It would appear that the OSF box is not correctly handling a request for
information about the DUA entity.
Can someone confirm this or ought I to take this to the DECnet/OSI for OSF
conference?
Rgds,
Graham
|
5985.3 | Yes there is a problem with DUA entity | TAEC::DEPAGNE | | Wed May 11 1994 10:31 | 16 |
|
Hi Graham,
I've also noticed the problem with my OSF machine.
You're right with your assumptions. Tracing the problem under MCC, it appears
that the Decnet osf cmip responder either sends back an empty message,
which is not well processed by MCC, or interrupts the communication or
something like that.
I don't know if the problem is known by decnet-osf team, anyway you
should take this problem to the DECnet/OSI for OSF conference, as you
proposed.
Vincent.
|
5985.4 | | WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Wed May 11 1994 10:59 | 3 |
| Thanks, Vincent
Entered as 1059 in the OSF DECnet/OSI conference
|
5985.5 | Can we turn off querying the DUA entity? | WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Thu May 19 1994 12:14 | 28 |
| Hi, again.
I have been following this in the NOTED::DNU_OSI conference and there seems to
be a feeling that there is a CML problem in DECnet/OSI for OSF.
The following suggestion has been made by one of the engineering group.
Any thoughts?
Rgds,
Graham
<<< NOTED::DISK$NOTES7:[NOTES$LIBRARY_7OF4]DNU_OSI.NOTE;2 >>>
-< DECnet/OSI for {ULTRIX,OSF/1} >-
================================================================================
Note 1059.8 Unable to use MCC (on VMS) to register a OSF/1 V2 machine with DECnet/OSI 8 of 9
NAC::HARRINGTON 10 lines 19-MAY-1994 09:34
--------------------------------------------------------------------------------
re .7
> I have asked the project manager responsible for this to get in touch with
> someone in your group to get a formal understanding of what is possible.
It might be worth the time to query the DECmcc folks, to see if the
request for a DUA entity can be supressed when the director tickles
the DECnet/OSI system. If this can be pulled off, then you'd have
a short term workaround.
Dan
|
5985.6 | See my reply to note 1059 | BIKINI::KRAUSE | CSC Network Management/Hubs | Fri May 20 1994 08:31 | 0 |
5985.7 | | WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Tue May 24 1994 19:00 | 15 |
| Thanks for that...
I don't know how my reply got into 1059!!!
I removed the DUA subclass and everything seemed to work.
I have now received a modified cml image which will hopefully remove
the necessity of getting rid of the DUA subclass but I just need to
test it.
Many thanks
Rgds,
Graham
|
5985.8 | | WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Wed May 25 1994 10:35 | 14 |
| And the story continues..
Using the modified dictionary (ie with DUA subclass removed) MCC hangs
when trying to look at multiple children.
For example if I try and look at NODE xxx OSI Transport Template *
it hangs after displaying around 10 of them.
It look slike there is a further interworking problem here.
Any thoughts?
Rgds,
Graham
|
5985.9 | 'known' problem | BIKINI::KRAUSE | CSC Network Management/Hubs | Fri May 27 1994 14:58 | 9 |
| > Using the modified dictionary (ie with DUA subclass removed) MCC hangs
> when trying to look at multiple children.
> For example if I try and look at NODE xxx OSI Transport Template *
> it hangs after displaying around 10 of them.
There is already an IPMT case for this problem (#MGO100563).
Guess who submitted it :-)
*Robert
|
5985.10 | | WELTM1::CRIDDLE | Graham Criddle, MCS Tech Consultant, 853-4015 | Mon Jun 06 1994 06:12 | 8 |
| Hi, Robert.
Any idea when this problem crept in or what it relates to?
Is there any indication as to when a fix should be available?
Rgds,
Graham
|
5985.11 | DNS entry disappears. | KETJE::PACCO | Horum omnium fortissimi sunt Belgae | Tue Jun 07 1994 12:13 | 8 |
| Sind we have OSF V2.0 systems in our network, none of these can
be registered any more.
We experience that every time we want to register, the node entry in
DNS disappears.
Regards,
Dominique
|
5985.12 | What happens ... | KETJE::PACCO | Horum omnium fortissimi sunt Belgae | Wed Jun 08 1994 14:48 | 65 |
| More information ...
First decnet_dns_register is executed.
Then manage register node <fullname> synonym <synonym>
Now the 'registration' hangs.
You get here what can be found in DNS about the (partially) registered entity:
dns> set dnscp confidence high
dns> sho obj .htl.m.nhtl01
SHOW
OBJECT bc:.htl.m.nhtl01
AT 08-JUN-1994:18:22:20
%X4D43435F02000100350000000000000000000 = %xa0820085bf817b8200659f83ff7f60000000
00010002000000000130000000000000010000000000000000000000000100000000000000000000
0030000000350000000101010000000e026000000000000001000000000000000000000000010000
00000000000000000000000000bf817a8200149f817e109ae1802edd82cd010000aa0004000504
%X4D43435F11000100000000000000000000000 = %xa08200bfbf817b359f83ff7f300000000001
00020000000001300000000000000100000000000000000000000000000000000000000000000000
000000bf817c820080a12a810105822208002b2c791e097537b22bf595001200010368746c01016d
01066e68746c30310000830100a28200508101398248040002000113000001000302000013010005
0200dec00100060a0049000caa000400233021040002000113000001000302000013010004000001
00060a0049000caa000400233020830100
%X4D43435F12000100350000000000000000000 = %xa0820071bf817b8200659f83ff7f60000000
00010002000000000130000000000000010000000000000000000000000100000000000000000000
0030000000350000000101010000000e026000000000000001000000000000000000000000010000
00000000000000000000000000bf817c820000
DNA$NodeSynonym = nhtl01
DNA$Towers (set) = :
Tower 1 Floor 1 = 01 13 (null)
Tower 1 Floor 2 = 03 00 13
Tower 1 Floor 3 = 05 de c0
Tower 1 Floor 4 = 06 49 00 0c aa 00 04 00 23 30 21
DNA$Towers (set) = :
Tower 1 Floor 1 = 01 13 (null)
Tower 1 Floor 2 = 03 00 13
Tower 1 Floor 3 = 04 (null)
Tower 1 Floor 4 = 06 49 00 0c aa 00 04 00 23 30 20
DNS$ACS (set) = :
...
MCC_Class = %x01000000
MCC_UID = %x368d84c6e582cd010000aa0004000504
MCC_Version = %xa68200089f817d0456010300
But then DECmcc registration HAS to be interrupted (registration otherwise
last for ever).
After the (interrupted) DECmcc registration (interrupted by ^C)
we have lost the object in DNS:
dns> set dnscp confidence high
dns> sho obj .dna_nodesynonym.nhtl01
SHOW
OBJECT bc:.dna_nodesynonym.nhtl01
AT 08-JUN-1994:19:05:24
Error on entity: bc:.dna_nodesynonym.nhtl01
Requested name does not exist
Function: dnsEnumAttr
dnsEnumAttr: partial results = bc:.dna_nodesynonym
dns>
|