Conference netcad::hub_mgnt

  More infos:

  Same Problem under HUBwatch for Digital Unix V4.0.1

  if you force the polling using polling refresh options
  the Decconcentrator900MX power led goes off others remain on.

for polling , use TEMIP !!!

Hubwatch is NOT a supervisor ; its a superb configuration tool 

for real-life monitoring use a SNMP supervisor or manager or director or 
you name it :

Manageworks/SNMP for small sites
Polycenter Netview for medium sites
TEMIP for large sites


salut Jean-Yves 

  use TEMIP !!!

   do you see the state of Leds with this ...


   jean-yves 
  

Hello Jean-Yves !!

i guess you want all operators buying heavy glasses from 
Afflelou 

i cant imagine an operator staying all day in front of 
the HUbwatch display trying to figure out which led went yellow
or red , with a hub fully populated with 8 repeaters  * 32 ports 


from an operator point of view it is better to have an alarm window
( a la Temip , or a la Netview , or whatever ) that will signal 
any event or trap  like  port state change 

now it would be great if we can have the details on how to setup 
the RMON A&E in repeater  Snmp agent , so that on port state change 
or on collision partitionning , would generate a trap 

PJ

>> i guess you want all operators buying heavy glasses from 
>> Afflelou 


 no i'm wondering that Hubwatch does'nt reflect the real state of materials...

  jean-yves

>now it would be great if we can have the details on how to setup
>the RMON A&E in repeater  Snmp agent , so that on port state change
>or on collision partitionning , would generate a trap
>
>PJ

Documentation of this feature is not a high priority.  I've attached an
older mail message that discusses this.
						-Shawn
--------------------------------------------------------------------------
From:	NETCAD::GALLAGHER    12-JUN-1995 13:43:47.92
To:	PRSSOS::PEYRACHE
CC:	NAC::FORREST,GALLAGHER
Subj:	re: Rmon and Repeater

Hi Jean-Yves,

RMON Alarm and Event functionality is in the MAM and several repeaters.
Unfortunately, there's no HUBwatch support yet.  RMON A&E is only useful
now for:

    1) default alarms and events which ship preconfigured in the
       products and cause SNMP traps to be sent when "interesting"
       things happen, and
    2) for those users who are willing to use a MIB browsers to manipulate
       the RMON objects.

Defaults A&E are documented in the product owner's manual.

If you want to learn to use a browser to manipulate the A&E objects I
recomend reading rfc1757.txt.  I'm also attaching a note I wrote when
we were thinking of including browser-type information into the owner's
manuals.  Unfortunately, I don't think it will be very useful to you.
It was written for the MAM.
						Regards,
						-Shawn

Default Alarms and Events
=========================

This section......


AlarmTable
----------

                       Sample           Startup Rising Falling Rising Falling
Index Int Variable      Type      Value  Alarm  Thresh Thresh  Event  Event   Owner   Status
----- --- ------------ ---------- ----- ------- ------ ------- ------ ------- ------- ------ 
  1   200 chasNumSlots deltaValue   4   falling   1      1       1      2     monitor valid
          Occupied


This example shows a default alarm table entry.  This entry
causes the object "chasNumSlotsOccupied" to be sampled every
2 seconds.  This example shows a hub containing 4 modules.
After a startup delay, if chasNumSlotsOccupied increases by
one (a module is inserted) a rising event is generated.
EventTable entry 1 describes the rising event.  If chasNumSlots-
Occupied decrease by one (a module is removed) a falling event 
is generated.  EventTable entry 2 describes the falling event.

 

EventTable
----------
                                                Last
Index  Description    Type          Community  TimeSent Owner    Status
-----  ------------   ------------  ---------  -------- -------  ------
  1    A module was   log-and-trap  public     6000     monitor  valid
       inserted. 
  2    A module was   log-and-trap  public     9000     monitor  valid
       removed.


EventTable entry 1 specifies that when then event is generated
a logTable entry is to be created and an SNMP Trap is to be
sent.  The trap is sent using the community string "public".
Traps are sent to all MAM traps sinks.  (Tom:  Pls refer the
user to the appropriate section on trap sinks.)


Modifying/Deleting/Creating Entries
====================================

Default alarms and events can be modified using an SNMP management
tool.  

For example, the community string used in trap messages for events 1 and
two can be changed using the following command: 

> set eventCommunity.1 "trap"
> set eventCommunity.2 "trap"

Default entries can also be deleted.  If you don't want to receive traps
or event logs when modules are inserted and removed then the alarm entry
should be deleted by setting the corresponding alarmStatus to invalid(4)
using the following command:

> set alarmStatus.1 4

To be neat, the event entries for this alarm could also be deleted as
follows:

> set eventStatus.1 4
> set eventStatus.2 4

Entries can also be created.  The alarmTable can be populated with as
many as 20 entries.  The eventTable can be populated with as many
as 32 entries.