[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

3558.0. "Another notification problem ?" by STKMCC::LUND (Niklas Lund) Thu Aug 13 1992 10:43

Hi !

We are using the following command procedure to test the data collector and
notification functions of DECmcc SSB 1.2 VMS.

$ i = 0
$ write sys$output "''f$time()'"
$loop:
$ i = i + 1
$ if i.gt.100 then goto exit
$ mcc_evc_send antik collector "node4 antik" "test" "Event:''i,''f$time()'" 'p1'
$ goto loop
$exit:
$ write sys$output "''f$time()'"
$ exit$ i = 0

We execute this procedure simultaneous in different processes using 
different severity values for P1. When the fourth process is started, we 
start loosing events and also getting some of them twice. 

The domain "colltst" contains one node4 antik and one collector named
"collector". The only notify request set up for this domain looks like this:  

Notify Domain colltst Entity List = (collector collector), Events = (any events)

No other notify request or alarm rules are active.

This has been reproduced on two different systems, one slow 3100 m38 and one
fast 4000-60. Both with 32 mb memory and VMS 5.5 / 5.5-1 and DECnet wave 2 / IV.

Regards Niklas.
T.RTitleUserPersonal
Name		DateLines
3558.1QAR 3356TOOK::MINTZErik Mintz, dtn 226-5033Thu Aug 13 1992 10:532
I have entered this note as QAR 3356.
3558.2clarify?TOOK::MCPHERSONLife is hard. Play short.Thu Aug 13 1992 10:5410
>This has been reproduced on two different systems, one slow 3100 m38 and one
>fast 4000-60. Both with 32 mb memory and VMS 5.5 / 5.5-1 and DECnet wave 2/IV.

    Are you running this from 4 different processes on one system to send
    to the event sink on the *same* system or to *another* system on the
    network running the sink?

    Are you seeing any unusual messages in the MCC_EVC_SINK.LOG file?

/doug
3558.3patch coming soonRAMPAL::S_KOHoot mon!Thu Aug 13 1992 11:526
    
    There is a patch being generated for this bug.  More information to
    come. 
    
    thanks,
    -s
3558.4STKMCC::LUNDNiklas LundThu Aug 13 1992 12:1525
RE .2

 >  Are you running this from 4 different processes on one system to send
 >  to the event sink on the *same* system or to *another* system on the
 >  network running the sink?

It happens both when sending and receiving on the same node and when we send
events to one sink from different nodes on the network.

When sending from different nodes the problems starts to appear when six nodes
are sending simultaneously. With this setup we also see a lot of
"network partner exited" messages.

[email protected] INDETERMINATE
13-AUG-1992 16:38:16.43
%SYSTEM-F-LINKEXIT, network partner exited
$

 >  Are you seeing any unusual messages in the MCC_EVC_SINK.LOG file?
 
No


Niklas
3558.5updateTOOK::S_KOHoot mon!Wed Sep 16 1992 08:4711
    
    hi,
    
    sorry for the long pause.  after some investigation, we did find a bug
    in the collection am.  note .3 references a different bug ;-}
    
    updated images are going to be generated for both cases.
    
    thanks for your interest and patience!!
    
    -s