[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

5116.0. "Exporting pauses with Thread termination" by USOPS::HARB () Tue May 25 1993 15:45

My background exporter process goes into a HIB state and never comes out after
the process encounters a CMA Exception:

%Thread 3 terminating with exception:
%CMA-F-EXCCOP, exception raised; VMS condition code follows
-SYSTEM-F-ACCVIO, access violation, reason mask=04, virtual address=00000000, PC=00128DB8, PSL=03C00004

I had a similar problem with the historian and that was due the number of record
requests that could be handled by any one recorder (500).  Does the exporter
have the same limitation?  If so, what is it?  Can it be increased?

I've checked all system parameters and process quotas and all look good.

I'm running DECmcc V1.2.1 on VMS V5.5-2


Thanks,

George
T.RTitleUserPersonal
Name		DateLines
5116.1TOOK::SHMUYLOVICHWed May 26 1993 11:3512
	
	This CMA exception does not mean that you exceeded the maximum number 
   of requests per one process. When this maximum is reached the Background
   process will terminate with the corresponding message ( some thing like
   "Maximum thread count is reached"). This is correct for both Historian and
   Exporter. Also both of them have the same maximum for active requests: 500,
   which can not be increased.

	SAm
5116.2Question remains?USOPS::HARBWed May 26 1993 12:5910
    Thanks for the reply.
    
    
    Ok, so what the the CMA exception noted in .0 mean?  Is there a memory
    leak?  And why doesn't the process terminate instead of hanging?
    
    
    Thanks,
    
    George
5116.3TOOK::SHMUYLOVICHWed May 26 1993 17:4826
>    Ok, so what the the CMA exception noted in .0 mean?

	Usually it means an access violation.

>    Is there a memory leak?

	Don't know. I don't see how memory leak can be related to this
        problem.

>    And why doesn't the process terminate instead of hanging?

	The Background Process (BG) is a multithreded process. The above 
	message tells that one of the threads is terminated but others are 
	still alive. They are competing for communication with another process 
	- Background Writer (BW) which is responsible for writing data into the 
	database. Only one thread at a time can communicate with the BW. 
	During this communication this thread is holding a lock indicating
	that BW is busy. If this thread terminates this lock is never released
	and eventually the BG process is hanging.

	I would recommend to upgrade to MCC V1.3. The Exporter in V1.3 has 
	some improvement in this area.

		Sam