Fellow list members,

A programmer in our site keeps submitting the following QUIZ code in an
attempt to extract information for some financial reconciliation task.
Each time, the job runs for many many hours, causing a bit of grief to the
rest of the staff/users.

Does anyone on this list have any suggestions (forgetting for a moment
about the obvious ones of using some other tool) for speeding the process
up?

So far the process has run for over 15 hours.  The first 'request'
extracted 960 records (I assume rather quickly) and it now seems to be
'stuck' in the second 'request'. Glance indicates that it is around the
half way point of processing the subfile from step one.  I should also
indicate that PCTRD0 is an image dataset with about 33 million entries.

I'd appreciate any constructive suggestions.

Regards

Paul Christidis

BTW the subfiles used are NOT indexed.

-----------------------------------------
SET REP LIM 2000
ACC *CHKQQQF1
SEL IF YR GT '1994'
set sub name chkqqqf2 keep
REP sum all
GO

ACC *CHKQQQF2 LINK ('0' + FIS) TO TRD0-FISCALNO OF PCTRD0
seL PCTRD0 IF TRD0-ID#4[2:3] = JRN AND &
              TRD0-ID#1 = PR AND TRD0-PACCOUNT[1:1] GE '9'
DEF HR NUM*10 = NCON(TRD0-UNITS/AMT[1:10]) * -1
DEF AM NUM*10 = NCON(TRD0-UNITS/AMT[11:10]) * -1
DEF PROJ CHAR*4 = TRD0-PACCOUNT[1:4]
DEF NEW-CHARGE CHAR*13 = TRD0-PACCOUNT OF PCTRD0[1:13]
SEL IF HR = HRS AND AM = AMT AND NEW-CHARGE[5:3] NE '999'
SET SUB NAME CHKQQQF3 KEEP
REP SUM PR PROJ FH GL YR HRS AMT NEW-CHARGE
GO


ACC *CHKQQQF3 LINK PROJ TO PROJ OF *CONVFILE OPT
SET PAG LEN 60
REP TAB 1 PR TAB 7 PROJ TAB 12 FH TAB 16 GL &
YR HRS AMT NEW-CHARGE CONV-FL GROUP
GO
EXIT
--------------------------------------------

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *