[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

627.0. "SHUTDOWN and CLEAR won't work" by SIEVAX::TMJ (Entropy eradicator) Wed Jan 16 1991 06:52

Is there something special about the behaviour of the SHUTDOWN verb?
The following are extracts from .MS and vector files:

MSL:

    DIRECTIVE SHUTDOWN = 50 :
	DIRECTIVE-TYPE = ACTION,
	DISPLAY        = TRUE,
	CATEGORIES     = ( CONFIGURATION ),

	RESPONSE Shutdown Success = 1 : 
	    SYMBOL = SHUTDOWN_SUCCESS,
	    TEXT   = "Shutdown successful.",
	END RESPONSE Shutdown Success ;

    END DIRECTIVE SHUTDOWN ;

Vector:

	MCC_DISPATCH_ENTRY -
		SHUTDOWN, -
		< MCC_ENTITY < ge, * > >, -
		NULL, -
		MCC_GE__SHUTDOWN

The results of executing the directive, using an AM built with the above, are:

>	$ mcc
>	DECmcc (T1.1.0)
>	
>	%MCC-I-PARTABBLT, parse tables built with: MCC_PTB version T1.1.0
>	MCC> shutdown ge test
>	User Command: shutdown ge test
>	
>	
>	*****************************************************
>	*     FCL PM Arguments before call_function:        *
>	*****************************************************
>	
>	VERB code:      50
>	PARTITION code: 10
>	AES dump of ENTITY IN:
>	
>	depth=1 class code= 107 instance = �
>	
>	ILV dump of IN_P: in_p is NULL
>	ILV dump of IN_Q: in_q is NULL
>	TIME SPEC is: 0, NOW
>	HANDLE STATE, before call: FIRST
>	
>	**********************************************
>	FCL PM Arguments on return from call_function:
>	
>	CVR returned:
>	%MCC-E-NOTFOUND,  unsupported combination of verb, entity, partition
>	
>	ILV dump of OUT_P:
>	
>	 NULL
>	
>	AES dump of ENTITY OUT:
>	
>	depth=1 class code= 107 instance = �
>	
>	HANDLE STATE, on return: FIRST
>	%MCC-E-NOTFOUND,  unsupported combination of verb, entity, partition
>	MCC>

The access module function (mcc_ge__shutdown) is not being called as it would
print a trace message.

A similar problem occurs with CLEAR.


And while on the subject of verbs, what is IGNITE (code 43) intended to do?!
T.RTitleUserPersonal
Name
DateLines
627.1PM's only go through function callsTOOK::HAOWed Jan 16 1991 09:3913
    The message that you're getting back means that the dispatch entry
    was not found.  Looking at the info printed out by the FCL, it looks
    correct according to your MSL and vector definitions.  
    
    One suggestion is to make sure that the Control FM can route 
    the Shutdown directive to the access module dispatch table.  The
    PMs can only make mcc_call_function calls, not mcc_call_access.
    Therefore, if your dispatch entries are in the access module table,
    and Control FM can't re-route the PM's call, you will see the
    "vea combo unsupported" message.
    
    Christine
    
627.2Function, not access, table; but why?SIEVAX::TMJEntropy eradicatorThu Jan 17 1991 04:4918
Thanks Christine.

I changed

    mcc_enr_enroll ( MCC_K_ACCESS, ... )

to

    mcc_enr_enroll ( MCC_K_FUNCTION, ... )

in the _init function and SHUTDOWN and CLEAR now get to my access module code.

But why? What is the difference between these verbs and, for example, SET and
SHOW? And are there other verbs that require this enrolling in the function
dispatch table? Is there any documentation relating to this?

						Tony M-J
627.3sorryTOOK::HAOThu Jan 17 1991 09:4310
    Hi Tony,
    
    Sorry, I don't know why the verbs are different.  The person who
    is responsible for the Control FM may be able to help you more, and
    unfortunately, I don't know who that is either.
    
    Jill, as moderator, would you have further info on persons to contact?
    
    Christine
    
627.4send request to registrar GOSTE::CALLANDERFri Jan 18 1991 12:3413
    
    Like any other module in the system the control FM must explicitly
    add in entry points for all VEP combinations. The system does not
    allow you to add in an entity point for a wildcarded verb. Because
    of this the control FM has a limited set of verbs that it can pass
    thru. 
    
    If you would like to see a new verb to be handled by the control
    FM please write up a quick request, including the verb and entity
    information and the need, and send it to Steve TOOK::WONG.
    
    Hope this helps.
    
627.5Now enrolled in both dispatch tablesSIEVAX::TMJEntropy eradicatorFri Jan 18 1991 13:3712
Thanks Jill, but I don't know if there's really a need to do that. What I've
ended up doing is enrolling my table in *both* the access and function dispatch
tables, so that *all* the verbs I'm using are recognised and dealt with by MCC.

It would be useful if this apparent difference in "behaviour" between some
verbs and others were documented in the MCC user doc. set. I (mistakenly)
thought that I could pick a verb with a name appropriate to my purposes from
the MCC_K_VERB_ set in MCC_VEA_DEF.H and deal with it in a similar fashion to
common verbs like SET, SHOW, etc.

						Tony M-J
627.6registrationGOSTE::CALLANDERFri Jan 18 1991 14:0213
    
    You would have found out when you attempted to register your management
    specification that things are a bit more strict then they first
    appear. Currently the DECmcc is in the process of changing hands
    so that the process can be better controlled and the "rules" of
    registration in MCC can be better understood. 
    
    As to documentation of what gets through control, I thought it was
    in the guide to writing an access module, if it is not then it should
    be qared.
    
    jill