| Hi Rex. I should've piped up earlier, but Mssrs Woodcock and Reilly seem to
have the right solution for you...
> The main issue here is for the short term (pre V2.0) how can the
> reporting/notification of alarms be consolidated to a single
> presentation ?
Per previous suggestions: you can pretty easily create alarm rules and use
the mcc_evc_send (Data Collector sample 'send' code) to generate a Data
Collector event based on the alarm data, then 'sink' the event(s) to the
central management system.
> Rather than have each director fire its own MAP & Notification window
> into the respective support group, does V1.2 have the functionality to
> RELIABLE report alarms to a SINGLE director, hence allowing a single
> presentation of alarms via MAP & Notification window.
Does this mean the the 'remote' DECmcc systems wouldn't really event need to
be workstations? I.e. you could just plunk down a box on the customer net
with just a terminal & maybe a high-speed modem for remote terminal access
(sorta like the Digital Service Processor model I remember from a few years
back...?). If so then your cost for the remote hardware goes way down...
If you don't want them to run the IMPM et al, then you don't need any fancy
graphics hardware on them...
> With V1.2 could a remote director upon detecting an alarm, fire a
> procedure over a wan link to update a notification on a central
> director which then relays that information via Iconic Map or
> Notification window to a support group ?
Yes. Absolutely. You just gotta know how to plug the 'gazoutas' into the
'gazintas'... ;^) Again, the most expedient way to accomplish this (i.e.
without coding up any tricky MMs) is to use the example data collector
application 'as is' from alarm fired procedures to sink events.
Additionally, you have the freedom with the Data Collector Event Name
argument to even include a keyword that indicates which site or LAN (or
whatever) the Data Collector event came from; then at the 'Central Grand
Kahuna DECmcc World Domination System' you can use notification services
filters to weed out display of events from different sites, if you need...
Neato, huh?
> The customer understands for control of LAN devices, a connection to
> the remote director is required. We simply want a single presentation
> of Notifications/Alarms not a display from each individual director.
DECmcc gives you the freedom to do either! Neato, too.
As a sidebar: market intelligence (?oxymoron? ;^) ) is sending us
the message that customers aren't really looking for 'freedom to
choose' as much as they are 'ease of use out of the box'. Believe
me, engineering is listening and as soon as they can dig out from
under the 1.2 deadlines I'm sure you'll start to see much more
happening in that area, too...
> Should these notifications proliferate through to a central display,
> how can they be governed ?
I don't understand what you mean by being "governed"... Are you familiar
with the features of the Notification Services PM (new for 1.2) ?
> We have a multiple operator environment, does one operator deleting the
> notification from their display, delete it for all others ?
Sorry, but you can't. Once a notification has been received, you can't
'undo' it. If you' re logging the events, then they're in the logfile
already. If, by 'delete the notification' you mean deleting if from the
Notification Services event window, then that ONLY affects the notfication
PM on the workstation that received the event.
> Is there a method of centrally controlling the notifications ?
>
Again, please be more specific about what you mean by 'controlling the
notifications'...
> Is any form of trouble ticketing available in the short-term, or are we
> talking V2.0 ?
None in the 1.2 timeframe. TNMP is working on something for the 2.0
timeframe. If you have near term requirements, we are working with
vendors who hhave applications in that area such as Target Systems'
"Target->Hotline" and UIS's "HelpLine". We havev established a good
relationship with Target Systems and have been able to demonstrate some
useful integration of their "Hotline" trouble ticketing application with
DECmcc. (See notes about trouble ticketing elsewhere in this conference...)
> I am looking for some technical answers to the above, but also
> practical ones as we a looking for a deliverable & supportable
> solution. Also to find out if anyone (customer or internal) is
> supporting a comparable environment with DECmcc BMS.
Good luck with ITT. Sounds promising. I think you can do what you want
using the methods proposed in this and previous replies.
regards
/doug
Rex
[1mHAYNES> [mlo
MCPHERSON logged out at 12-MAR-1992 08:28:11.96
Local -011- Session 1 disconnected from TOOK
Local> lo
Local -020- Logged out port 16 on server LKG_DIALIN_703
X�����A��u��P��L�J�'�?�HT.��UM^A:�aw[�V�'g�h➦��Ru��9��IW�ä
=7
�HZ�/`��b5E�T�*��T�k}�&7Zz��x��0�K�n"�{���6 cQw{P���$4_d~�W�>���lp<�%Nn�<�������d��}�W"�
LEB�d!�(#�i)�t�CTV-�SL��r��?�p�Sg>�>�]}�
� >�imp�?���r�dP��
|
|
DOUG et all....
Thanks for your replies. As ever the funds available to the customer
has shrunk the state of this particular bid.......but this is how it
now shapes.
We are looking to use DECmcc/Ultrix and the UDMT product, plus some
'enhancements' from the OSEC in Manchester (UK).
We will have initially 3 directors, 1 central to receive the hub of
notifications which will be passed using the 2 remote directors. To do
this remote alarms are passing information via the data collector am on
the central director.
Several UDMT's remote agents running on Personnel DECstations will sit
on each LAN polling local ethernet devices for availability. We are
hoping for modifications to the 'standard' UDMT remote agent here so a
poll of a terminal server not only tells you the device is powered but
software operational as well.
The OSEC are looking into offering SNANCP info from a SNAG, DTE info
from an X25 G/W and maybe info from an ICL G/W to the users with
modified UDMT or DECmcc modules.
We are also investigating passing info from DCMx to DECmcc and DECmcc
to CHOx to facilitate further needs.
If succesful the customer is going to grow this configuration
throughout all of his UK network.
Regards,
Rex
|