[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

4104.0. "ALARMS evaluation is not adequate to analyze problems" by CUJO::HILL (Dan Hill-Net.Mgt.-Customer Resident) Wed Nov 18 1992 00:31

If you need this functionality, please respond to this note:

*** DECmcc does not perform adequate DELTA evaluations via Alarm Rules. ***

Let me clarify.  Let's say that I want to check for a broadcast storm on
the network.  I can use the CHIPCOM MIB variable (SNMP attribute)
onlineEnetStatsBcastReceivedOks.  I'll poll the Chipcom hub for this variable
every 20 seconds.  Subtracting the current poll's value from the previous
gives me a difference, or delta.  If that delta is greater than 700 (35/sec),
there's a good chance I have a broadcast storm on my hands.

The CHANGE_OF alarm rule format allows me to detect a change, but doesn't
tell me how much of a change.
-----------------------------------------------------------------------------
I need to be able to alarm on this DELTA if it is above a certain threshhold.
-----------------------------------------------------------------------------
This is a MAJOR HOLE in DECmcc's functionality.  I would like to see ALARMS
extended to support complex rules, something like an alarms rule evaluator.
                                                     ------ ---- ---------
The PA_FM does some of this, but only for selected attributes.  You can't do
it on the fly for a given MIB variable or other attribute.
    
    This would be a big step towards solving some of the major network
    problems that can occur.  I was caught badly off guard when I found
    this functionality missing. 
    -Dan

T.R	Title	User	Personal Name	Date	Lines
4104.1	Also EMF_REQ?	TOOK::MINTZ	LKG2-2 near pole X3, cube 6072, dtn 226-5033	`Wed Nov 18 1992 00:40`	2
	This sounds like a topic that should (also?) go in NOTED::EMF_REQ
4104.2	Already discussed in note 1336	BERN01::GMUER		`Wed Nov 18 1992 08:28`	6
	This item was already discussed in note 1336. There were also some suggestions to extend the CHANGE_OF rule format to solve this requirement. But nothing happened until today ! Edgar
4104.3		TOOK::MINTZ	LKG2-2 near pole X3, cube 6072, dtn 226-5033	`Wed Nov 18 1992 08:42`	12
	Engineers are often interested in discussing new ideas in this conference. Sometimes, suggestions are incorporated in code as a result. However, significant changes to functionality require formal planning. The people who have the greatest influence on the order of requirements in the priority list do not follow this conference. If you have strong requirements (and can justify them with customer needs) then make sure product management is aware of those requirements either through the NOTED::EMF_REQ conference, or directly.
4104.4	Already posted in EMF_REQ	CUJO::HILL	Dan Hill-Net.Mgt.-Customer Resident	`Wed Nov 18 1992 15:34`	4
	I should have stated this in .0, but I simultaneously posted this note in both conferences. -Dan
4104.5	Disappointed	CUJO::HILL	Dan Hill-Net.Mgt.-Customer Resident	`Wed Nov 18 1992 15:49`	7
	I just read 1336.* . I am very disappointed that this functionality was never implemented. Aside from "COUNT OF's" in the STATISTICS partition, we're out of luck. This should be "fixed" ASAP. How can we help? -DH
4104.6	Re .3: Make it simpler !	BERN01::GMUER		`Thu Nov 19 1992 10:42`	25
	re .3 > Engineers are often interested in discussing new ideas in this conference. > Sometimes, suggestions are incorporated in code as a result. > However, significant changes to functionality require formal planning. OK, but I always thought that the notes conferences are the way to exchange ideas in an informal way and I still hope that we could tell the engineers and the product management what is going on out in the market. Comparing this delta-rule requirement to other over-sophisticated features of DECmcc I think it is peanuts. > The people who have the greatest influence on the order > of requirements in the priority list do not follow this conference. Perhaps they should ! This conference reflects the situation at the customers. DECmcc is a good product, but we have to make it more competitive by decreasing the complexity of the usage. Comparing to the products of the competitors it is still too complicated. Edgar
4104.7		TOOK::MINTZ	LKG2-2 near pole X3, cube 6072, dtn 226-5033	`Thu Nov 19 1992 11:35`	12
	Sorry, I guess I over stated something in .3 We certainly want to encourage an exchange of ideas here. I was just trying to point out that when something is really important, make sure it gets tracked by using the formal mechanism (which in the case of requirements, means using the notes conference that product management has chosen to set up exclusively for requirements). Thank you for your thoughts. -- Erik