[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference ilbbak::ibi_focus

Title:FOCUS, from INFORMATION BUILDERS
Moderator:ZAYIUS::BROUILLETTE
Created:Thu Feb 19 1987
Last Modified:Mon May 05 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:615
Total number of notes:1779

587.0. "TABLE FILEs not returning control." by PAKORA::JHANLON () Thu Aug 25 1994 08:55

Hi, 

I have a problem with a fex which has been working and has suddenly ceased to do
so !

I have two "Table file..."s with "on table hold"s, one after the other, with the
second one reading the datafile produced by the first. I then wish to match the
second hold file to yet another data file. However, after the second "On Table
Hold", the fex does not get any further. I have checked the process, and it is
still doing something using CPU time although IO count does not increase. It
appears to be looping.
 Both hold files are very large, but we have other fexes doing "on table hold"
on the same file to retrieve similar amounts of data and they work.

This only happens when run on the DECScheduler (Mode is 'Other') or running in
Batch mode. Executing the fex interactively seems to work fine. 


The Focus code and output looks as follows:


TABLE FILE FILE1
BLA BLA BLA
ON TABLE HOLD AS 'FILE2'
END

TABLE FILE FILE2
BLA BLA BLA
ON TABLE HOLD AS 'FILE3'
END

MATCH FILE3
BLA
RUN
FILE FILE4
BLA
AFTER MATCH..etc


but the Match is never started.


......

 MERGE:   RECORDS=   136947 LINES=    93023
 MERGE:   RECORDS=   139908 LINES=    95238
 MERGE:   RECORDS=   143115 LINES=    97449
etc.
 MERGE:   RECORDS=   176067 LINES=   118656
 NUMBER OF RECORDS IN TABLE=   176067  LINES=   118652
 HOLDING...

 MERGE:   RECORDS=     3448 LINES=     3448
 MERGE:   RECORDS=     6896 LINES=     6896
etc.
 MERGE:   RECORDS=   118652 LINES=   118652
 NUMBER OF RECORDS IN TABLE=   118652  LINES=   118652
 (BEFORE TOTAL TESTS)

 HOLDING...


And then .....nothing.

If anyone can help me with this, or knows somenone who can, I will be eternally
grateful !!
Thanks,

    Jill.
T.RTitleUserPersonal
Name
DateLines
587.1Suggestions..PEKING::MCSHANEGDeceptively Co-operativeSun Aug 28 1994 16:2420
    Jill,
    
    	A few thoughts that might help...excuse me if you've tried this
    stuff already...
    
    You say this problem only occurs when in batch/sched. Could the problem
    be something to do with missing files or directories that are not
    available via logicals etc when running in batch ?. Try assigning the
    dir's containing the .MAS's and .DAT/.FOC files to the FOC$DIRn
    logicals as part of the procedure.
    
    Is there no error in a log file ?. Try turning -&ECHO=ON . You should
    then be able to pick up on the failure point/error.
    
    Have you used $ ASSIGN/USER 'F$LOGICAL("TT")' SYS$INPUT As part of the
    .COM that runs in batch ?
    
    Gary
    
    
587.2Try splitting the input upBRAT::WEHIKE::ClarkWed Sep 07 1994 12:4713
We had that happen at some point and I am sorry I can't be sure I  remember the problem.  I seem to think 
that it had to do with the size of the output file being larger than the memory.  We found that the request to 
create a long record ( 256 characters) x 1,000,000 lines resulted in a process that created a hold file, but 
the file wasn't there when we looked for it.    So we had the same view in the log.  It created something, said 
holding, but never "finished".

If all of -1's suggestions have been tried and aren't applicable, try splitting the data up into 2 logical pieces ( 
by alphabet, or numerical first digit or something) and retry the fex thru the scheduler.  I AM puzzled that 
interactively it works, but the internal memory assignments might be different or the scratch areas assigned 
might be diferent when running in different modes......

Good luck.
Debi