[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

1919.0. "MM Development Questions -- the With Qualifier" by MOLAR::ROBERTS (Keith Roberts - DECmcc Toolkit Team) Wed Dec 11 1991 09:06

This is a new Topic Note continuted from note

		1862.7

to discuss Management Module Development with regards to With Clause Processing

RE: 1862.10

>    Re : 1862.7
>    
>    Keith,
>    
>    Am I right in understanding this reply as saying that the WITH
>    qualifier is *not* passed down to the AM ? If this is the case, this is
>    in violation with the SRM. Page 62 states that :
>    
>    "The WITH qualifier is associated with the in_entity parameter of the
>    VEP tuple. Specifically,  the WITH qualifier is placed at the
>    appropriate level in the entity hierarchy in the In_Entity parameter
>    
>    (...)
>    
>    One WITH qualifier may be used at the lowest level of the entity
>    hierarchy for MCC V1.0"
>    
>    Furthermore, table 5.1 states that the with qualifier may be handled by
>    the PM, the IM *or* the target MM.
>    
>    Please clarify this point, as my customer is basing his AM design on
>    this , and one of his customer's requirements is that a minimum amount
>    of filtering is supported.
>    
>    I believe I can make him accept that only on WITH clause is supported
>    and only for base datatypes, but no support at all will really put us
>    in trouble (to say it mildly!).
>    
>    This is the only way I'd have of supporting some kind of CMIS-like
>    scoping and filtering.
>
>    Philippe (CCIIS1::ROGGEBAND)

  Philippe,

  Yes, the SRM does talk about the With Qualifier.  The SRM is sometimes
  ahead of the actual implementation of DECmcc...this is one of those times.

  It was the intention to put the With Qualifier into the In-Entity Specifier,
  but this hasn't happened yet, and will not (to the best of my knowledge)
  happen in V1.2

  You are correct about the With Qualifier not being passed to your MM.
  It is only processed by the PM's.  Your MM returns all the partition
  attributes, and the PM does the filtering.

  Keith
T.RTitleUserPersonal
Name
DateLines
1919.1CCIIS1::ROGGEBAND_ �hili��e _Wed Dec 11 1991 09:1210
    Keith,
    
    My my, well we are in trouble now.... (At least, I am). How would you
    go about implementing something which looks loke filtering ? Any
    suggestions will be appreciated.
    
    Are there any plans for this to be implemented in the not-too-far-away
    future ?
    
    Philippe. 
1919.2The PM does the filter - so it does get doneMOLAR::ROBERTSKeith Roberts - DECmcc Toolkit TeamWed Dec 11 1991 09:4821
RE: .1

>    My my, well we are in trouble now.... (At least, I am). How would you
>    go about implementing something which looks loke filtering ? Any
>    suggestions will be appreciated.

  Philippe, the PM does do the With Qualifier processing, so filtering
  is done -- you realize that, right?  Your MM keeps returning all the data
  that is requested by the Verb, In-Entity and Partition arguments, and
  the PM ignores data which the With Qualifier would reject.

  But you need the With Qualifier passed to your MM - Right ?  What does
  your MM need use it for -- do you need to send the filter to the Entity ?

>    Are there any plans for this to be implemented in the not-too-far-away
>    future ?

  That, I do not know -- I'm sorry.  I'm not sure where it appears on the
  list of things for the next release either.

  Keith
1919.3CCIIS1::ROGGEBAND_ �hili��e _Wed Dec 11 1991 10:558
    Keith,
    
    What the customer wants is to be able to send out the filter to the
    agent to avoid flooding the network with unnecessary data. As the
    filtering capability is a part of the CMIP protocol, it is not very
    efficient to let the PM do all the work.
    
    Philippe.
1919.4I figured that was itMOLAR::ROBERTSKeith Roberts - DECmcc Toolkit TeamWed Dec 11 1991 13:4520
RE: .3

> What the customer wants is to be able to send out the filter to the
> agent to avoid flooding the network with unnecessary data.

I figured that was it, but I just wanted to make sure.  I know of no way
to get the With Qualifier to the MM.  Until an engineered solution has been
ECO'd into DECmcc, I'm sorry to say, the MM must return all the data for the
PM to filter  8(

I hope when the time comes, we (the toolkit) have the opportunity to get
some library routines in place to make the job of processing the filter
easier for the MM developer.  If a CMIP-like format is used, you can forward
it to the Entity ... but other MM developers will have to process the With
Qualifier manually.

/keith



1919.5closeTOOK::MATTHEWSThu Dec 12 1991 14:2411
    There are some basic issues within MCC as to how to implement
    filteringg
    and to define when to pass the scope and filter and where it
    should be applied. I put a note in this conference outlining the
    issues about 6 months back. Please read it to understand the
    issues.
    
    Keith - I think TNMP already has filter routines  they are the same
    ones as defined for Forwarding discriminators
    
    wally
1919.6We'll have to leverage off of the TNMP routinesMOLAR::ROBERTSKeith Roberts - DECmcc Toolkit TeamThu Dec 12 1991 17:0412
Thanks Wally -- I'll read over the note you mentioned (have any idea which
one it is?)

> I think TNMP already has filter routines  they are the same
> ones as defined for Forwarding discriminators

This is what we want to hear!  Code Reuse - Great!

We should leverage off of any tools already available, and eliminate
as much redundant code as possible - I like it 8)

/keith
1919.7TOOK::STRUTTManagement - the one word oxymoronSun Dec 15 1991 20:197
    Not mentioned in previous replies, referring to the ability for the
    "PM to do the filtering", is that the current scheme only works on SHOW 
    directives.
    
    Clearly a solution needs to be implemented in order to allow other (non 
    SHOW) operations, as well as the desire to save network bandwidth (an 
    important consideration in itself).
1919.8Code reuse is great !TAEC::LAVILLATMon Dec 16 1991 04:1848
Re .6:
>
>> I think TNMP already has filter routines  they are the same
>> ones as defined for Forwarding discriminators
>
>This is what we want to hear!  Code Reuse - Great!
>
>We should leverage off of any tools already available, and eliminate
>as much redundant code as possible - I like it 8)
>

Code reuse is great but...

I am in charge of the filtering routines for TNMP (now called TeMIP).

We develop in fact 4 kinds of functionality. They are all based on OSI filters
(CMISFilter type) :

	1/ A MOTIF based window interface to set and modify filters (seen as
	object attributes within TeMIP).

	2/ A set of library routines to store/retrieve/update/delete/list/...
	pre-defined filters (using the MIR as repository )

	3/ A set of manipulation routine to create/build/read/decompose filters
	(to manipulate filters with as less as possible ILV knowledge ...)

	4/ A set of routine to validate and evaluate filters.

The reuseability problem is that our filters are used exclusively to filter
OSI events and so filtering routines will take in input a filter and an OSI
event and tell you if they match or not. Only part 2/ and 3/ can be reused
as such, 1/ can be extended without too much problem, but 4/ (which is the
most interesting part) would need a lot of improvements to be used for
directive filtering. 

Another point is that we have the requirement to be able to filter at least 25
events per second, which is not much to access the dictionary and be fully
generic...

You may also check with the Common Agent team, since they have implemented
a fully generic set of filtering routines based also on CMIS filters, but ...
using AVL as data encoding instead of ILV...

Regards.

Pierre.

1919.9There's always a 'But...'NANOVX::ROBERTSKeith Roberts - DECmcc Toolkit TeamMon Dec 16 1991 13:209
RE: .8

Thanks for the information Pierre.  I don't know when or how the With
Qualifier stuff will be done -- but I hope its based on standards, such as
the CMIS filters, and made simple enough for MM developers to easily 
incorporate into their code.

/keith

1919.10with clause in FCLTOOK::CALLANDERMCC = My Constant CompanionTue Jan 14 1992 17:364
    what kind of filtering routines are you interested in, the FCL also has
    some (since it is the one supporting with right now) that handles
    expression evaluation for the with clause.