T.R | Title | User | Personal Name | Date | Lines |
---|
3014.1 | | METSYS::POAD | http://www.digital.com/info/edi | Mon Feb 17 1997 13:33 | 5 |
| The V2.1 Cockpit kit should work with V2.1C Servers.
There is a problem with ObjectBroker where V2.5A on the server has
problems with V2.6 on the PC (and client I think) such that multiple
server process were started which weren't shutdown.
|
3014.2 | It WONT work | METSYS::HELLIAR | http://samedi.reo.dec.com/ | Mon Mar 03 1997 09:26 | 13 |
| Mike,
The Unix client wont work as its OBB version is 2.5B or above whilst
the server is 2.5A. OBB 2.5A does not understand the 'new' handles
generated by OBB 2.5B and above.
Note this change in handle was done at 2.5B so is only a problem where
clients are 2.5B or above and server is 2.5A. Problem does not exist if
server is at 2.5B or above.
Mark should have a matrix of the supported combinations.
Graham
|
3014.3 | The matrix location | SYSTEM::HELLIAR | http://samedi.reo.dec.com/ | Fri Mar 07 1997 11:14 | 3 |
| Mark's support article containing the matrix:
http://samedi.reo.dec.com/deug/secure/support.html
|
3014.4 | EDI 3.1/OBB 2.7 -> 2.1C/OBB 2.5A? | CSC32::R_GOLLEHON | | Mon Mar 10 1997 19:26 | 24 |
| Hello,
>Mark's support article containing the matrix:
>http://samedi.reo.dec.com/deug/secure/support.html
I noticed that 2.7 OBB doesn't appear in the matrix. Does this imply
that it doesn't work or hasn't been tested?
I have a customer who has a 3.1 Digital Unix client which he is trying
to get to work with a 2.1C server...right now he has OBB 2.7 on the
client and 2.5A on the server....the matrix suggests that this will not
work...is this correct? (Customer had 3.1A client, but he is in the
process of downgrading to 3.1 base, so I can't play with it quite yet).
In the customer's current config (3.1A client), I have been able to get
trade track to work, but post/fetch fails with an OBB_REG_KEYNOTFND
before the server even gets touched.
Thoughts?
Thanks,
-Robert
|
3014.5 | IT WONT WORK | METSYS::HELLIAR | http://samedi.reo.dec.com/ | Tue Mar 11 1997 08:59 | 7 |
| I have a customer who has a 3.1 Digital Unix client which he is trying
to get to work with a 2.1C server...right now he has OBB 2.7 on the
client and 2.5A on the server....the matrix suggests that this will not
work...is this correct?
>> Also see .2 where I said it wont work due to OBB changing its
>> internal handles between 2.5A and 2.5B
|
3014.6 | | METSYS::THOMPSON | | Tue Mar 11 1997 09:10 | 7 |
| Robert,
A version of that matrix with OBB V2.7 is under preparation.
Should be available shortly. Treat OBB V2.6 and V2.7 as pretty much the
same.
M
|
3014.7 | working, thanks | CSC32::R_GOLLEHON | | Tue Mar 11 1997 14:52 | 8 |
| My apologies Graham...should have looked a bit more closely...after
looking for a while we reached the same conclusion and consequently
the customer is now using OBB 2.5A on the client side and it works like
a champ.
Thanks,
-Robert
|
3014.8 | 2.1D, OBB 2.7, OVMS 6.2, DECNET-OSI 6.3 | NNTPD::"[email protected]" | Pertti L�hteenm�ki | Tue Jun 03 1997 08:50 | 36 |
| Hi,
I need your help to find out should this configuration work at the first
place.
I was optimistic and
- upgraded D/E 1.3A to D/E 2.1D and installed
- a new version of OBB 2.7 and
- upgraded Rdb 4.2 to Rdb 6.1 and
- DECforms 1.4 to 2.1B.
All was fine, but not ObjectBroker. In this system were Decnet-OSI v6.3
installed.
The error I've got from tracing ObjectBroker is:
Transport Status: %OBB-E-INV_COMMFAIL, Transport level communicaton failed.
Operation Status: %OBB-E-INV_COMMFAIL, Transport level communicaton failed.
%DECEDI-E-ORBERROR, ObjectBroker Failure
In DECEDI.ERR file I found the following error messages:
-OBB-E-INV_TRANSERR, Transport error
%SYSTEM-F-LINKABORT, network partner aborted logical link'.
Module : DECEDI_API_POST, %OBB-E-INV_COMMFAIL, Transport level communication
failed.
Do you have an idea which version of Decnet-OSI and ObjectBroker I should have
or
what other matters should I have thought here?
Thank you for your help!
Regards,
Pertti
[Posted by WWW Notes gateway]
|
3014.9 | | METSYS::THOMPSON | | Tue Jun 03 1997 11:26 | 12 |
| Hi,
OBB V2.7 has been tested on DECnet-plus (V7.1) and that works just fine.
Never tested on OpenVMS V6.2 or DECnet/OSI V6.3 but I know of no reason
why that should not work.
Did you read the OBB V.27 Release Notes? There is an important configuration
change mentioned in there that has been discussed elsewhere in this conference.
Did you run DECEDI$CONFIGURE_OBB?
Mark
|
3014.10 | DECEDI$CONFIGURE_ACAS? | NNTPD::"[email protected]" | Pertti L�hteenm�ki | Tue Jun 03 1997 13:48 | 21 |
| Hi Mark,
Yes I did. Is there any difference to use decedi$configure_acas not ..obb?
ObjectBroker asked when installing if to use Decnet as transport which was
right in
that node. D/E client and server were in the same node. I looked that
everything
was fine in 'inter edit config' and 'register client applic' in 'configure
nodes'
and I alse did build/replace cache.
When client and server are in the same node. What should be typed in
'configure nodes' in 'objectbroker required' yes/no?. I did try both and there
were no difference, error message was the same. I does some configuration for
ObjectBroker though when yes is typed there.
Regards,
Pertti
[Posted by WWW Notes gateway]
|
3014.11 | another approach | HLSW01::DEWIJK | GJ from the Dutchlands | Tue Jun 03 1997 14:09 | 9 |
| Hi,
the SYSTEM-F-LINKABORT message could indicate that the receiving
system, your own system, aborts the incoming call.
I suggest to have a look in the accounting file if there is any reason
mentioned what happened; eg captive-flags, diskquota, nonetwork-flag
A guess,
GJ
|
3014.12 | | METSYS::THOMPSON | | Tue Jun 03 1997 19:08 | 22 |
|
>> Yes I did.
Did you do the "modify configuration" ?
> Is there any difference to use decedi$configure_acas not ..obb?
That's gone into at length in other notes. Basically decedi$configure_acas.com
can prompt Objectbroker to crash leaving a damaged registry that prevents
all Objectbroker applications working. It's an Objectbroker bug.
decedi$configure_obb.com incorporates a re-order of two lines that
seem to prevent the crash.
Not all systems fail, i.e. I expect this is due to someone running with
low quota's or priv's. If you ran decedi$cofigure_Acas and it didn't
give you any problems then -don't worry.
>When client and server are in the same node. What should be typed in
>'configure nodes' in 'objectbroker required' yes/no?. I did try both and there
Use Yes to Objectbroker required.
|