[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference azur::mcc

Title:DECmcc user notes file. Does not replace IPMT.
Notice:Use IPMT for problems. Newsletter location in note 6187
Moderator:TAEC::BEROUD
Created:Mon Aug 21 1989
Last Modified:Wed Jun 04 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:6497
Total number of notes:27359

4447.0. "Problem with IP Poller, alarms and event sink" by RDGENG::PRATT () Tue Jan 26 1993 05:50

T1.3.1
VMS V5.5-2

2 problems..

1) With T1.3 I was successfully using the IP Poller, and had an alarm rule which
fired when the IP poller detected a change in reachability of a IP host. Since
upgrading to T1.3.1 this alarm rule no longer fires, although I get the
notification event okay. I assume this is something to do with the fact that the
IP poller now uses Notification Events instead of Configuration events (see note
4354.9).

The alarm rule is ...

Domain MCCULT_NS:.internet Rule IP_Reachability_down
AT 26-JAN-1993 10:17:45 Characteristics

Examination of attributes shows:
                  Alarm Fired Procedure = DISK$USER01:[MCC_ALARMS]MCC_ALARMS_WIN
                                          DOW_ALARM.COM;3
              Alarm Exception Procedure = DISK$USER01:[MCC_ALARMS]MCC_ALARMS_WIN
                                          DOW_EXCEPTION.COM;4
                            Description = "IP reachability down"
                               Category = "IP"
                            Batch Queue = "mcc_alarms"
                 Alarm Fired Parameters = ""
                             Expression = (occurs (SNMP *  IP Reachability
                                          DOWN))
                               Severity = Major
                         Probable Cause = Unknown

Can anyone suggest why this alarm doesn't work anymore?

2) The second problem is that I have a batch job that disables/re-enables all of
my alarm rules every midnight, and occasionally the above alarm rule fails to 
enable itself (this is not related to problem (1) above)...


Domain MCCULT_NS:.internet Rule IP_Reachability_down
AT 26-JAN-1993 10:07:14 Status

Examination of attributes shows:
                                  State = Disabled
                               Substate = Time specification has expired
                           Disable Time = 26-JAN-1993 00:01:11.48
                Time of Last Evaluation = 26-JAN-1993 00:01:10.86
              Result of Last Evaluation = Error
                        Error Condition = "The Event Sink is running, but not
                                          communicating; cannot restart Sink. "
                           Error Entity = SNMP *
                       Current Severity = Indeterminate

What event sink is it looking for? because I can re-enable the alarm manually 
without any problems - and it doesn't happen every time.

Ian
T.RTitleUserPersonal
Name
DateLines
4447.1You'll have to re-create the RuleMOLAR::ROBERTSKeith Roberts - Network Management ApplicationsTue Jan 26 1993 08:3312
> fired when the IP poller detected a change in reachability of a IP host. Since
> upgrading to T1.3.1 this alarm rule no longer fires, although I get the
> notification event okay. I assume this is something to do with the fact that the
> IP poller now uses Notification Events instead of Configuration events (see note
> 4354.9).

  The data stored 'inside' the Rule is no longer valid because of the change
  made by SNMP; changing the IP Poller Event from Configuration to Notification.

  The fix - delete and recreate the rule.

  /keith
4447.2No luckRDGENG::PRATTTue Jan 26 1993 10:575
Keith,

I've deleted and recreated the rule, but it still doesn't fire.

Ian
4447.3is it in the domainGOSTE::CALLANDERTue Jan 26 1993 15:095
    the occurs rule in V1.3 now expands the wildcard based on the
    contents of the domain.  Please make sure that the snmp entity
    is in the domain, and that the registered name to internet ids
    (name and address) are what you are expecting.
    
4447.4Fix to Problem (1) will be in V1.3 SSBTRM::KWAKFri Jan 29 1993 16:3115
    
    RE: Problem (1) in .0
    
    The Alarms FM (T1.3.0 and T1.3.1) has a problem in processing Wildcarded 
    OCCURS rules when DECdns is used. The domain membership filtering causes 
    a problem.
    
    The fix will be available for V1.3 SSB.
    
    If you need to use the wildcarded OCCURS rules AND are using DECdns
    (both VMS and Ultrix), and cannot wait until V1.3 SSB, please send me 
    a mail (TOOK::KWAK).
    
    Sorry for inconvenience.
    William
4447.5Correction to .4TOOK::KWAKSat Jan 30 1993 14:2410
    
    Correction to .4
    
    You can have a problem with the Alarms FM if you use the following:
        1. Wildcarded OCCURS rules AND
        2. DECdns, AND
        3. SNMP entities in the OCCURS rule expression 
    	   (Other entities will not cause problems).
    
    William