Conference azur::mcc

    Ian,

    Please, describe your problem in more details.

    Did you have an Exporter running successfully in T1.2.4? in V1.1?

>    I have checked to see if any data had been exported  to the RDB file.
>    It had'nt!! 

	Do you use a new Rdb file or this file has been in use since V1.1?
	How did you check data in the RDB file? Are there any other data in 
        this file? Are there any tables created in this file?

>	It also looks like there's no recording going on either.
      
      What is a result of Show Exporting?

	Is this a past or future Exporting Request?


	SAm
    

    Yes I had an Exporter process running in both V1.1 and T1.2.4. When I
    stopped the T1.2.4 process and restarted it,then I got the RWMBX
    problem.
    
    I used  the old V1.1 RDB file. I backed up the old V1.1 VMS system from
    MCCULT to the new MCCDUF system and installed the T1.2.4 software onto
    the new system MCCDUF. I then started the T1.2.4 Exporter background
    process sucessfully and MCC> EXPORT NODE4 . When I found that it was'nt
    Exporting data to the RDB file,I stopped the Background process and 
    eventually restarted it. That is when the process went into RWMBX
    state.
    
    I checked the RDB file for data using both DECdecision and SQL. I was
    running the Exporter and then checking to see if any data had been 
    placed in the RDB file.
    
    My main concern at present , is the RWMBX state.
    
    Ian 

   Not exactly the same situation, but the symptoms are similar.

This problem is not yet resolved, but our Customer still has a need.

Is there any progress on this "RWMBX" issue, or anything we can do to 
help get the problem resolved?


       Richard.

            <<< NOTED::DISK$NOTES4:[NOTES$LIBRARY_4OF5]MCC.NOTE;1 >>>
                 -< DECmcc Network Management Product Family >-
================================================================================
Note 1182.17   Trouble to get the Exporter background job started.      17 of 17
BAHTAT::BOND                                         33 lines  31-OCT-1991 04:49
                       -< Another RWMBX - any progress? >-
--------------------------------------------------------------------------------
    Yes - we too are having some problems with the exporter background
    process sticking in RWMBX state - but the subprocess then seems to be
    doing nothing at all.   Suspend or delete exporting commands eventually
    give "request queue full - check background process" and you have to
    kill it off in the end.  Has any progress been made with this whole
    RWMBX issue?  I know Sam in a previous note implied this was 'normal'
    but it seems to me that when the exporter is working normally, it
    HIBernates; when it gets stuck in this RWMBX state and never comes out
    it is dead!  Perhaps somebody could explain how it should all work?
    
    At our customer site, we were trying to export NODE, LINE and CIRCUIT
    counters from 54 DECrouter nodes with 2 to 4 lines on each.  This gave
    398 exporting jobs to 1 database.  We found that the background process
    would never properly start; it would compute for 10 minutes or so
    gradually increasing its virtual address space and working set to some
    45000 and 16000 pages (WSEXTENT) respectively and then finally stick in
    RWMBX after about another 10 to 15 minutes.  Approaching this time, it
    would begin to flit into RWMBX more and more until finally it never
    came out.
    
    We then reduced the number of nodes being monitored to a Core set of 22
    which only have 2 lines on them giving 110 exporting jobs.  With this
    reduced load, the background process seems to start ok - it still flits
    into RWMBX wait when the working set has got to 16000 pages but it
    always gets out again.  It still takes about 15 minutes to get to this
    state and then happily remains in HIB state except when it is working.
    
    So, it seems to me that the exporter seems to be running out of
    something when it is asked to do many more jobs and then gets stuck in
    RWMBX.  Anybody any idea what it is?  I have checked FILLM (500),
    BYTLIM (200000) and there is plenty of non-paged pool.
    
    The system is a VS3100 model 76 with 32Mb memory.  

    Hi Richard,
    I have eventually got the Export background process to run and actually
    Export data!!
    I have stopped and deleted RECORD NODE4 and EXPORT NODE4 and the
    HISTORY and EXPORT queues. I've also deleted the V1.1 RDB and SNP
    files. Have restarted all above again successfully and using 
    DECdecision can actually see the exported data. 
    
    Are you using the MCC V1.1 RDB files? If so (from Notes 2247.2) you
    can't export data using MCC T1.2.4 into RDB file created by MCC V1.1.
    
    I hope that this helps!!!
    
    Kind Regards
    Ian
     

    
        Hello Ian,
    
       Thanks for the response. Our problem is not one of "getting
    exporting started" , in fact we have been exporting successfully for
    several weeks. The problem is "how much" exporting can be achieved.
    We cannot export on the number of nodes the Customer wants without
    hitting the "RWMBX" problem.
    
      regards,
    
            Richard.