JCLcnv1demo.htm - DEMO conversions, sample JCL, scripts, executions JCLcnv2real.htm - comprehensive instructions for REAL conversions JCLcnv3aids.htm - conversion AIDS (cross-refs,tips,mass changes,etc) *THIS DOC* JCLcnv4gdg.htm - GDG file handler from UV Software DATAcnv1.htm - Data file conversion - Original Documentation DATAcnv2.htm - Data File Conversion - Comprehensive & Complete
| Part_1 | Cross-References available |
| xrefall - script to run all cross-ref reports | |
| xkshfile1 - list all datafiles in each JCL/ksh/script | |
| xkshfile2 - shows all JCL/ksh/scripts using each data-file | |
| xkshparm1 - list all SYSIN/PARM modules in each JCL/ksh/script | |
| xkshparm2 - shows all JCL/ksh/scripts referencing each SYSIN/PARM | |
| xkshparmsd1 - alternate crossref if parms in multiple subdirs | |
| xkshparmsd2 - alternate crossref if parms in multiple subdirs | |
| xjclproc1 - list all PROCs executed in each JCL/ksh/script | |
| xjclproc2 - shows all JCL/ksh/scripts executing each PROC | |
| xkshprog1 - list all programs executed in each JCL/ksh/script | |
| xkshprog2 - shows all JCL/ksh/scripts executing each program | |
| xkshprog2a - alternate crossref combining all COBOLs | |
| xcobcopy1 - list all 'copy'books used in each program | |
| xcobcopy2 - shows all programs using each 'copy'book | |
| xcobcall1 - list all 'call'ed programs in each program | |
| xcobcall2 - shows all programs 'call'ing each program | |
| xcobfile2 - crossref all PROGRAMS using each external-file | |
| xcobsql1 - list all SQL Includes in any 1 PROGRAM | |
| xcobsql2 - crossref all PROGRAMS using any 1 SQL Include | |
| cobfiles5A - COBOL file report | |
| Statistics (table-summaries) available | |
| statmvsjcl1 - statistics for mainframe JCL | |
| - count programs,PROCs,top-nodes,files used in all JCLs | |
| statksh1 - statistics for Korn shell scripts | |
| - count programs,exportfile/exportgen,filenames in Ksh scripts | |
| statcbl1 - statistics for COBOL programs | |
| - count copybooks, called programs,& file ORG type | |
| mvsfiles5A - script to determine essential Input files for ALL JCL/scripts | |
| - extracts all data files from jcl2 (JCL procs expanded) | |
| - report xmvs/mvsfiles5 '*' flags essential input files | |
| mvsfiles51 - determine essential input files for 1 JCL vs All JCLs | |
| - invaluable during test/debug |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| Part_2 | Test/Debug Tips & Techniques |
| Test Environment, RUNLIBS & RUNDATA superdirs defined in your profile | |
| - allows different programmers to have their own set of Libraries & Data | |
| See JCL converted to scripts & executed in JCLcnv1demo.htm | |
| Use these tips for testing the Demo JCL/scripts or for your own JCL/scripts | |
| Converting Your DATA files - brief review of DATAcnv1.htm. | |
| run 'testdatainit' before JCL/scripts | |
| - to clear output files & make it easier to see outputs of current test | |
| - use 'joblog' to capture the log from your JCL/script | |
| Test/Debug for difficult JCL/scripts | |
| - save all data1/* in datasave/ & load data1/ with files for difficult job | |
| Iterations of test/investigate/modify as required | |
| - Animator/Debugger for Micro Focus COBOL | |
| Check results of test/debug, investigate output files (use uvhd if packed) | |
| Print-outs using 'uvlp' scripts assist test/debug | |
| Modifying no of gnerations in GDG control file & reloading Indexed file | |
| GDG files & step Restart | |
| Activating 'console-logging' vs 'job-logging' | |
| File Comparison for files with packed/binary &/or no LineFeeds | |
| - unix 'diff' does not work for these types of files | |
| - UV Software provides the 'uvcmp' utilities | |
| - uvcopy jobs uvcmp1,2,3 & several scripts uvcmpFA1,uvcmpFE1,etc | |
| to make the uvcopy jobs easier to run. | |
| - uvcmp prints mismatched record pairs in vertical hexadecimal | |
| flagging differences with '*'s, see sample report comparing | |
| 2 generations of gl.account.master_000001 & _000002 |
| Part_3 | Updating control-files to improve JCL conversions |
| Control-Files supply data-file-info that may be missing in JCL | |
| (record-sizes,file-types,GDG generations,etc) | |
| - required for script conversions of SORTs,IDCAMs,IEBGNENERs,etc | |
| capturing COBOL info for JCL conversion (record-sizes, file-types) | |
| control files created by 1st conversion (in subdir ctl/...) | |
| - datajcl51, datajcl52 extracted from the JCL | |
| JCL converter script 'jcl2ksh51A' does everything | |
| - calls 'jcldata51A' to create datajcl51 & datajcl52 | |
| & load into Indexed file datactl53I.dat/.idx for JCL converter | |
| - calls 'jclxx51' for JCL conversion to Korn shell scripts | |
| - updates datactl53I with datafile info from COBOL conversions | |
| using LISTCAT mainframe reports to improve JCL conversion | |
| - transfer, translate, store multiple reports in subdir cat0/... | |
| LISTCAT report sample (mock-up for VU demo files) | |
| converting Excel spreadsheets of datafile info to control files | |
| Re-Converting JCL with LISTCAT, datamisc,& dataxls | |
| - using 'alldiff2' script to see differences made by re-conversion | |
| - confirm chages as expected & no unintended changes | |
| Modifying JCL converter 'options' | |
| - example to change fixed length data files to Line Sequential | |
| - if you have very few files with packed/binary fields | |
| Determine Essential INPUT files for any 1 JCL/script | |
| - script 'mvsfiles51' selects essential input files | |
| - ignoring output files & intermediate created/used within JCL |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| Part_4 | Checking conversion results, Searching for potential problems. |
| Tools to search directories for patterns with qualifiers. | |
| 'grep' is the 1st choice to search for 1 pattern. | |
| 'selectf1' will search for up to 3 patterns with occurrence counts | |
| =, >, or < specified values. Default is AND, may specify OR. | |
| See example to search for SORT's with 'INCLUDE COND' or 'OMIT COND' | |
| 'scan1d' searches a directory for all lines matching 1 pattern | |
| 'scan2d' searches a directory for all lines matching any pattern | |
| in a table of pre-edited patterns | |
| 'scan3d' selects lines from a starting pattern to an ending pattern, | |
| qualified by a 3rd pattern present anywhere between start & end. |
| Part_5 | Mass changes to all files in directories of JCL &/or COBOL. |
| - summary of tools available | |
| Mass change example using 'sed' & script 'runsed1' | |
| - uncomment all '#rm' file deletes after test/debug | |
| - example using 'runsed1' to prompt for sed 's' substitute commands | |
| & apply while copying all files in directory to a 2nd directory | |
| Mass change example using sed, runsed1,& runsed2 | |
| - change all 'ksh' to 'ksh93', for clients that do not want | |
| to upgrade their current ksh (ksh88) to ksh93 (best for JCL/scripts) | |
| Mass change using uvcopy job 'rep1d' | |
| rep1d - copies all files from 1 dir to a 2nd dir | |
| - prompts for 1 pattern to be replaced by a 2nd pattern | |
| - optionally qualified by 2 other patterns (present or absent) | |
| Mass change using uvcopy job 'rep2d' | |
| rep2d - copies all files from 1 dir to a 2nd dir | |
| - searching/replacing depending on a pre-edited table | |
| of search/replace/qualify patterns | |
| rep3d - replace blocks of code in all files in a directory | |
| - blocks identified by a start pattern, an end pattern,& | |
| an optional qualifier pattern between start & end patterns | |
| uvfix1 - easy way to fix various problems in datafiles & scripts. | |
| - allows you to apply the powerful uvcopy instructions | |
| - translates, convert data field type (packed/binary/zoned) | |
| - example given to correct EBCDIC zoned signs to unix standards | |
| uvfixA - same as uvfix1, but processes all files in a directory | |
| while copying to the tmp1/ directory |
| Part_6 | Optional Conversion Enhancements |
| Directories required (Libraries & Data) | |
| JCL conversion OPTIONS control file (jclunixop51) | |
| - modify options to improve conversions depending on site data types | |
| datafiles containing many packed/binary fields or mostly character | |
| Add data file info onto JCL/script file definitions (as #comments) | |
| - appends Record-Sizes & Packed/Binary | |
| - examples of before/after adding datafileinfo as #comments | |
| ctl/datactl53I - control file info for JCL conversion |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| Part_7 | Test/Verify Critical Features in Converted JCL/scripts |
| Test 'joblog2' run JCL/scripts serially | |
| Prove joblog2 serial scheduler works using jar100.ksh | |
| - insert prompt/reply, run job from 2 screens, note delay | |
| - answer prompt on 1st job to complete & allow 2nd job to run | |
| Test automatic GDG generation# re-sequence at 900000 | |
| - gdgrest1 resets gdg#s for 'ONE' file automatically, whenever gdg# | |
| 900000 exceeded. 'gdgreset1' is called by function 'jobend51' which was | |
| inserted by the JCL converter at the end of all JCL/scripts. | |
| Test batch job to re-sequence gdg#s > specified threshold | |
| - gdgreset2 will reset GDG#s for 'ALL' files whose GDG#s exceed a | |
| specified threshold (suggested default 500000). You would run this | |
| batch job when no JCL/scripts are being executed. |
Might add 'IP# replace' to part 6 here in JCLcnv3aids.doc from JCL2cnvreal.doc#1E5 - 1E7 & 2A1
Might add jclunixop51 listing somewhere in this doc ?
select2d, xkshprog2a, xrefdrop1 - noted on page '1A1'
chkrcs1, chkrcs2, chkrcs2A - also noted in Part_6 contents
uvcopy chkrcs1,fild1=jcl3,filo2=rpts/chkrcs1.rpt ================================================ - check record sizes on uvsort/uvcp in JCL/scripts - see MVSJCL.doc or www.uvsoftware.ca/mvsjcl.htm
uvcopy chkrcs2,fild1=jcl3,filo1=rpts/chkrcs2_fj,filo2=rpts/chkrcs2_jf ===================================================================== - check record sizes on uvsort/uvcp in JCL/scripts - report datafilenames for rcs=99999 from saved SORTIN exports
chkrcs2A jcl3 <-- execute uvcopy job chkrcs2 + sorts =============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 1A1. | xrefall - script to run all cross-ref reports |
| 1B1. | xkshfile1 - list all datafiles in each JCL/ksh/script |
| 1B2. | xkshfile2 - shows all JCL/ksh/scripts using each data-file |
| 1C1. | xkshparm1 - list all SYSIN/PARM modules in each JCL/ksh/script |
| 1C2. | xkshparm2 - shows all JCL/ksh/scripts referencing each SYSIN/PARM |
| - alternate for sites with duplicate parmnames in multiple subdirs | |
| xkshparmsd1 - alternate parms in each JCL/script | |
| xkshparmsd1 - alternate JCL/scripts referencing each parm | |
| 1D1. | xjclproc1 - list all PROCs executed in each JCL/ksh/script |
| 1D2. | xjclproc2 - shows all JCL/ksh/scripts executing each PROC |
| 1E1. | xkshprog1 - list all programs executed in each JCL/ksh/script |
| 1E2. | xkshprog2 - shows all JCL/ksh/scripts executing each program |
| 1E3. | xkshprog2a - same as xkshprog2, but consolidates all COBOL programs |
| to highlight the various utilities & 3rd party software |
| 1F1. | xcobcopy1 - list all 'copy'books used in each program |
| 1F2. | xcobcopy2 - shows all programs using each 'copy'book |
| xcobcall1 - list all 'called'programs used in each program | |
| xcobcall2 - shows all programs calling each 'called'program | |
| 1G1. | xcobfile2 - crossref all PROGRAMS using each external-file |
| 1H1. | cobfil51 - COBOL Files Report |
| (not exactly a cross-reference, but a very useful report) | |
| - information extracted from all COBOL programs in directory | |
| - filenames, Input/Output, Org, Access, record size, copybook | |
| - all programmers should have a copy before starting test/debug |
| 1I1. | Statistics (table-summaries) available |
| statmvsjcl1 - statistics for mainframe JCL | |
| - count programs,PROCs,top-nodes,files used in all JCLs | |
| statksh1 - statistics for Korn shell scripts | |
| - count programs,exportfile/exportgen,filenames in Ksh scripts | |
| statcbl1 - statistics for COBOL programs | |
| - count copybooks, called programs,& file ORG type |
| 1J1. | jcl3_programs - table summary all programs executed by converted scripts |
| - highlights mainframe utilities that may need replacement | |
| 1J2. | jcl3_exportfiles - summary of exportfiles(non-GDG) & exportgens(GDG files) |
| - count exportfile/exportgen gen#s & subdirs | |
| 1J3. | jcl3_parms - table summary of all SYSIN/PARM libarary modules |
| 1J4. | jcl3_filenames - table summary of all filenames in JCL/ksh/scripts |
| - may be a long report (vs above short reports) |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 1K1. | cbls_copybooks - table summary counts of all copybooks used by all programs |
| 1K2. | cbls_ORGtypes - table summary of file ORGANIZATION types used |
| (SEQUENTIAL,LINE SEQUENTIAL,INDEXED,DYNAMIC,RANDOM,etc) |
| 1L1. | mvsfiles5A - script to trace data files in MVS JCL |
| - extracts all data files from subdir jcl2 (JCL procs expanded) | |
| - identifies the essential INPUT files to execute JCL/scripts | |
| - INVALUABLE during test/debug | |
| mvsfiles51 - for 1 JCL at a time (vs all JCLs in jcl2/*) |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl --> $HOME/testlibs (alias cdl='cd $HOME/testlibs)
#3. xrefall cbls jcl3 ksh <-- create all cross-ref reports
===================== COBOL/copybooks & JCL/scripts
#4. cobfiles5A cbls cpys maps <-- create COBOL 'FILES' report
=========================
| Mar31/2017 |
|
uvcopy select2d,fild1=xref,fild2=xrefmissing,uop=a0b1,arg1='*' ============================================================== - create 'missing modules' reports for parms, programs, copybooks
| Note |
|
uvcopy xrefdrop1,fili1=xref/xkshprog2a,filo1=xref/xkshprog2a_sum ================================================================ - drops all but 1st & last lines of utilities xref to create short report showing the total number of each utility with a few JCL names where used
#4. uvcopy statksh1 <-- create stat reports from Korn shell scripts
===============
#5. uvcopy statcbl1 <-- create stat reports from COBOL programs
===============
| Note |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
`xrefall`above creates reports `xkshparm1` & `xkshparm2`. which is sufficient for most sites. `xrefallparm`creates additional crossrefs as follows:
| xkshparmjps1/2 |
|
| xkshparmjp1/2 |
|
| xkshparmp1/2 |
|
| xkshparmp1 |
|
| xkshparmp2 |
|
'xkshparmp1' was the main objective here, to get a list of all parms used by proograms such as FASTLOAD which required special conversions (convert DECIMAL to varchar).
#1. mkdir step1 step2 step3 step4
=============================
#2. xrefallparm jcl3
================
xrefalltdsql is for sites that maintain a separate parms subdir for systems such as TeraData (BTEQ,FASTLOAD,MLOAD). The SQL parameters (load/unload tables) are maintained in separate parms subdir 'tdsqls'. xrefalltdsql creates following reports:
| xkshtdsqlp1 |
|
| xkshtdsql2 |
|
#3. xrefalltdsql step4
==================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xkshfile1 ** list all DATAFILES used in each ksh SCRIPT ** Keyword=exportfile:exportgen Exclude=autoload:JTMP:SYOT Include=~~ Skip=~~ page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a16b50c1j4w2q0i7 =======================================================2011/11/05_16:35:15
eztlist.ksh $RUNLIBS/parms/eztlist1 ___________ data1/ar.customer.eztlist ____________3 data1/ar.customer.master
ftpput1.ksh FTPPUT01
jar100.ksh data1/ar.customer.master ___________2 data1/ar.customer.nameadrs.list100
jar200.ksh data1/ar.customer.master.indexed __________ data1/ar.sales.items ___________3 data1/ar.sales.list
jgl100.ksh ACCTMAS ___________2 ACTLIST
jgl200.ksh $RUNLIBS/parms/pgl200s1 __________ GLMSNEW __________ GLMSOLD __________ GLTRANS __________ SORTOUT ___________6 data1/gl.account.tran1
jgl300.ksh ACNTMAS ___________2 SYSUT1
jgl320.ksh ACNTMAS ___________2 SYSUT1
qjtlist.ksh $RUNLIBS/parms/qjtlist1 ___________ data1/ar.custmas.tellist ____________3 data1/ar.customer.master
**TotalFiles: 9, TotalWords: 24, TotalRefs: 24
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xkshfile2 ** crossref to show all ksh SCRIPTs using each DATAFILE ** Keyword=exportfile:exportgen Exclude=autoload:JTMP:SYOT Include=~~ Skip=~~ page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a40b18c3j4w2q0i7 =======================================================2011/11/05_16:35:15
$RUNLIBS/parms/eztlist1 eztlist.ksh
$RUNLIBS/parms/pgl200s1 jgl200.ksh
$RUNLIBS/parms/qjtlist1 qjtlist.ksh
ACCTMAS jgl100.ksh
ACNTMAS_2 jgl300.ksh jgl320.ksh
ACTLIST jgl100.ksh
FTPPUT01 ftpput1.ksh
GLMSNEW jgl200.ksh
GLMSOLD jgl200.ksh
GLTRANS jgl200.ksh
SORTOUT jgl200.ksh
SYSUT1_2 jgl300.ksh jgl320.ksh
data1/ar.custmas.tellist qjtlist.ksh
data1/ar.customer.eztlist eztlist.ksh
data1/ar.customer.master_3 eztlist.ksh jar100.ksh qjtlist.ksh
data1/ar.customer.master.indexed jar200.ksh
data1/ar.customer.nameadrs.list100 jar100.ksh
data1/ar.sales.items jar200.ksh
data1/ar.sales.list jar200.ksh
data1/gl.account.tran1 jgl200.ksh
**TotalWords: 20, TotalFiles=24, TotalRefs: 24
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xkshparm1 ** crossref to show all ksh SCRIPTS referencing each PARM ** Keyword=parms Exclude=~~ Include=exportfile Skip=~~ page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a16b16c4d1j4l1w1q0i7 =======================================================2011/11/05_16:35:15
eztlist.ksh eztlist1
jgl200.ksh pgl200s1
qjtlist.ksh qjtlist1
**TotalFiles: 3, TotalWords: 3, TotalRefs: 3
xkshparm2 ** crossref to show all ksh SCRIPTS referencing each PARM ** Keyword=parms Exclude=~~ Include=exportfile Skip=~~ page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a16b16c4d1j4l1w1q0i7 =======================================================2011/11/05_16:35:15
eztlist1 eztlist.ksh
pgl200s1 jgl200.ksh
qjtlist1 qjtlist.ksh
**TotalWords: 3, TotalFiles=3, TotalRefs: 3
xkshparmsd1 & xkshparmsd2 are for sites that had duplicate parmnames in different PDS libraries on the mainframe, so on unix we cannot combin all parms into 1 subdir parms/... We must store in multple subdirs
parms <-- subdir for all parms when no duplicate names
parmsds <-- superdir for multiple subdirs when parmnames not unique :-----pgapo_abe_ekstada_vlk <-- multi subdirs : :--------------------artgrp : :--------------------artgw <-- parm modules :-----pgapo_abe_katalog_vlk : :--------------------adwh400 : :--------------------artgw <-- may be duplicate names
See details for parms in multiple subdirs at JCLcnv2real.htm#1E1 - 1E8.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#xjclproc1 ** crossref all MVS JOBnames for each EXEC PROCname ** #Keyword=EXEC Exclude=PGM Include=~~ Skip=~~ page# 1 #Directory=/home/mvstest1/testlibs1/jcl1 #Options=q1a16b16c4e0d0g0h1j0l1p0s2w1a16b16c4j15w1q0i7 #=======================================================2016/02/15_12:38:24
jgl100.jcl pgl100
jgl200.jcl pgl200
#**TotalFiles: 2, TotalWords: 2, TotalRefs: 2
#xjclproc2 ** crossref all MVS JOBnames for each EXEC PROCname ** #Keyword=EXEC Exclude=PGM Include=~~ Skip=~~ page# 1 #Directory=/home/mvstest1/testlibs1/jcl1 Options=q1a16b16c4e0d0g0h1j0l1p0s2w1a16b16c4j15w1q0i7 #======================================================2016/02/15_12:38:24
pgl100 jgl100.jcl
pgl200 jgl200.jcl
#**TotalWords: 2, TotalFiles=2, TotalRefs: 2 #**Total Missing Procs = 0 **
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xkshprog1 ** list all PROGRAMs executed in each ksh SCRIPT ** Keyword=progid:program Exclude=~~ Include=~~ Skip= page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a18b16c4d8j4l1q0i7 =======================================================2011/11/05_16:35:15
eztlist.ksh eztpa00
ftpput1.ksh ftp
jar100.ksh car100
jar200.ksh_2 car200 sort
jgl100.ksh cgl100
jgl200.ksh_2 cgl200 sort
jgl300.ksh_2 idcams iebgener
jgl320.ksh_2 idcams iebgener
mailsmtp.ksh iebgener
qjtlist.ksh quikjob
**TotalFiles: 10, TotalWords: 14, TotalRefs: 14
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xkshprog2 ** crossref to show all ksh SCRIPTS executing each PROGRAM ** Keyword=progid:program Exclude=~~ Include=~~ Skip= page# 1 Directory=/home/userxx/testlibs/jcl3 Options=q1a16b16c4e0d0j0l0s2w1a16b16c4d8j4l1q0i7 =======================================================2011/11/05_16:35:15
car100 jar100.ksh
car200 jar200.ksh
cgl100 jgl100.ksh
cgl200 jgl200.ksh
*eztpa00 eztlist.ksh
*ftp ftpput1.ksh
*idcams_2 jgl300.ksh jgl320.ksh
*iebgener_3 jgl300.ksh jgl320.ksh mailsmtp.ksh
*quikjob qjtlist.ksh
*sort_2 jar200.ksh jgl200.ksh
**TotalWords: 10, TotalFiles=14, TotalRefs: 14
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#xkshprog2a ** crossref to show all ksh SCRIPTS executing each PROGRAM ** #Keyword=progid:program Exclude=~~ Include=~~ Skip= page# 1 #Directory=/home/mvstest1/testlibs1/jcl3 Options=q1a16b16c4e0d0g0h1j0l1p0s2w1a16b16c4d8j4l1q0i7 #======================================================2016/02/15_12:38:24
*cobol jar100.ksh jar200.ksh jgl100.ksh jgl200.ksh ______6 jgl230.ksh_2
eztpa00 eztlist.ksh
ftp ftpput1.ksh
icetool icetool3.ksh
idcams jgl320.ksh
iebgener_2 jgl320.ksh mailsmtp.ksh
quikjob qjtlist.ksh
sort_3 jar200.ksh jgl200.ksh jgl230.ksh
#**TotalWords: 8, TotalFiles=15, TotalRefs: 16
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xcobcopy1 ** list all COPYBOOKS in each PROGRAM ** Keyword=copy:include Exclude=unix Include=.cpy Skip= page# 1 Directory=/home/userxx/testlibs/cbls Options=q1a16b16c4e0d0j0l0s2w1a16b16c4e3l1q0i7 =======================================================2011/11/05_16:35:14
car100.cbl custmas.cpy
car200.cbl_4 custmas.cpy saledtl.cpy sdline.cpy stline.cpy
**TotalFiles: 2, TotalWords: 5, TotalRefs: 5
xcobcopy2 ** crossref all PROGRAMS using each COPYBOOK ** Keyword=copy:include Exclude=unix Include=.cpy Skip= page# 1 Directory=/home/userxx/testlibs/cbls Options=q1a16b16c4e0d0j0l0s2w1a16b16c4e3l1q0i7 =======================================================2011/11/05_16:35:14
custmas.cpy_2 car100.cbl car200.cbl
saledtl.cpy car200.cbl
sdline.cpy car200.cbl
stline.cpy car200.cbl
**TotalWords: 4, TotalFiles=5, TotalRefs: 5
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
xcobfile2 ** crossref all PROGRAMS using each external-file ** Keyword=external Exclude=~~ Include=~~ Skip= page# 1 Directory=/home/userxx/testlibs/cbls Options=q1a16b16c4e0d0j0l0s2w1a16b16c4e1l1q0i7 =======================================================2011/11/05_16:35:14
acctmas cgl100.cbl
actlist cgl100.cbl
custmas_2 car100.cbl car200.cbl
glmsnew cgl200.cbl
glmsold cgl200.cbl
gltrans cgl200.cbl
nalist car100.cbl
saledtl car200.cbl
salelst car200.cbl
**TotalWords: 9, TotalFiles=10, TotalRefs: 10
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
cobfiles5A cbls cpys maps <-- create COBOL 'FILES' report =========================
cobfil51 ** COBOL Files Report ** Dir=cbls 2011/10/23_16:39:38 progname.cbl DDname OAM open recsz copybook.cpy FDname / Key lines ================================================================================
car100.cbl custmas SS_ I___ 256 custmas.cpy custmas
car100.cbl nalist L__ O___ 120 nalist
52
car200.cbl saledtl SS_ I___ 64 saledtl.cpy saledtl
car200.cbl custmas IR_ I___ 256 custmas.cpy custmas
car200.cbl key-> cm-cust
car200.cbl salelst L__ O___ 120 sdline.cpy salelst
65
cgl100.cbl acctmas SS_ I___ 128 acctmas
cgl100.cbl actlist L__ O___ 120 actlist
57
cgl200.cbl glmsold SS_ I___ 128 glmsold
cgl200.cbl glmsnew SS_ O___ 128 glmsnew
cgl200.cbl gltrans SS_ I___ 80 gltrans
65
** Note re COBOL Files Report **
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| statmvsjcl1 |
|
| statksh1 |
|
| statcbl1 |
|
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl --> $HOME/testlibs (alias cdl='cd $HOME/testlibs)
#3. mkdir stats <-- make subdir for reports
===========
#4. uvcopy statmvsjcl1 <-- create stat reports from MVS JCL
==================
#5. uvcopy statksh1 <-- create stat reports from Korn shell scripts
===============
#6. uvcopy statcbl1 <-- create stat reports from COBOL programs
===============
#7. vi stats/* <-- view stat reports
==========
#8. uvlpd12 stats <-- print all reports
============= OR print only desired reports
- most stats/reports short
- BUT, filename reports could be very long
#ra. uvlp12 stats/jcl3_programs <-- print table summary of executed programs
========================== - 1 of shorter useful reports
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
statksh1 2011/10/28_13:16:05 Summary table of PROGIDs in jcl3
tbl#0001 tblt1f4 c0(32)
line# count % 1strec# PROGID
1 1 7 19 car100
2 1 7 47 car200
3 1 7 25 cgl100
4 1 7 53 cgl200
5 1 7 19 eztpa00
6 1 7 18 ftp
7 2 14 44 idcams
8 3 21 22 iebgener
9 1 7 19 quikjob
10 2 14 19 sort
14*100 *TOTAL*
statksh1 2011/10/28_13:16:05 Summary exportfile/exportgen subdirs in jcl3
tbl#0002 tblt2f4 e100(32)
line# count % 1strec# exportfile/gen [0/+1] subdir/
1 12 30 22 exportfile $JTMP/
2 3 7 31 exportfile $RUNLIBS/
3 3 7 30 exportfile $SYOT/
4 10 25 21 exportfile data1/
5 5 12 24 exportgen0 0 data1/
6 6 15 47 exportgen1 +1 $JGDG/
39*100 *TOTAL*
statksh1 2011/10/28_13:16:05 Summary table of parms/modules in jcl3
tbl#0004 tblt4f4 d900(30)
line# count % 1strec# parms/module
1 1 33 23 eztlist1
2 1 33 31 pgl200s1
3 1 33 23 qjtlist1
3*100 *TOTAL*
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
statksh1 2011/10/28_13:16:05 Summary table of data1/filenames in jcl3
tbl#0005 tblt5f7 d300(48)
line# count % 1strec# subdir/filename
1 1 2 28 $JGDG/data1/gl.account.acntlist_
2 3 7 47 $JGDG/data1/gl.account.master_
3 2 5 29 $JGDG/data1/gl.account.trans_
4 1 2 29 $JTMP/${JSTEP}_${PROGID}_FTPTMP2
5 1 2 20 $JTMP/${JSTEP}_${PROGID}_INPUT1
6 3 7 23 $JTMP/${JSTEP}_${PROGID}_SYSIN
7 1 2 21 $JTMP/${JSTEP}_${PROGID}_SYSUT1
8 4 10 25 $JTMP/__tempglm
9 2 5 22 $JTMP/__tempsls
10 1 2 23 $RUNLIBS/parms/eztlist1
11 1 2 31 $RUNLIBS/parms/pgl200s1
12 1 2 23 $RUNLIBS/parms/qjtlist1
13 2 5 30 $SYOT/${JSTEP}_SYSOUT
14 1 2 20 $SYOT/${JSTEP}_SYSUT2_B_SMTP
15 1 2 22 data1/ar.custmas.tellist
16 1 2 22 data1/ar.customer.eztlist
17 3 7 21 data1/ar.customer.master
18 1 2 50 data1/ar.customer.master.indexed
19 1 2 22 data1/ar.customer.nameadrs.list100
20 1 2 21 data1/ar.sales.items
21 1 2 51 data1/ar.sales.list
22 4 10 24 data1/gl.account.master_
23 1 2 28 data1/gl.account.tran1
24 1 2 28 data1/gl.account.trans_
39*100 *TOTAL*
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
statcbl1 2011/10/24_09:36:59 Summary of copybooks in cbls
tbl#0001 tblt1f4 d0(32)
line# count % 1strec# copybook
1 2 15 27 custmas.cpy
2 1 7 25 saledtl.cpy
3 1 7 29 sdline.cpy
4 1 7 30 stline.cpy
5 4 30 65 unixproc1.cpy
6 4 30 46 unixwork1.cpy
13*100 *TOTAL*
statcbl1 2011/10/24_09:36:59 Summary of ORGANIZATION types in cbls
tbl#0003 tblt3f4 d0(32)
line# count % 1strec# organization
1 1 10 15 indexed
2 3 30 20 line sequential
3 6 60 10 record sequential
10*100 *TOTAL*
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Scripts 'mvsfiles51/mvsfiles5A/mvsfiles5B' create "data file trace" reports from jcl2 (MVS JCLs with procs expanded). There are 7 reports created mvsfiles1-mvsfiles7 but the most useful report might be 'xmvs/mvsfiles5' which '*' flags the INPUT files actually needed to run any 1 job or all jobs in subdir jcl2/*.
This report eliminates duplicates & intermediate files (outputs for later inputs) & identifies the essential files required to be be transfered from the mainframe. There are 3 scripts as follows:
| mvsfiles51 |
|
| mvsfiles5B |
|
| mvsfiles5A |
|
For this demo, we have run script mvsfiles5A based on 2 jobs (cgl100.jcl & cgl230.jcl) listed on pages '2C1' & '2E1'.
| Note |
|
You have already run 'jcl2ksh51A' (on page '3F3'), but you must now run 'cobfiles5A' before you run mvsfiles51 or mvsfiles5A. cobfiles5A creates Indexed files from COBOL copybooks & programs to supply record-sizes & packed field indicators.
#1. jcl2ksh51A all <-- converts all jcl0/* --> jcl1 --> jcl2 --> jcl3
============== - creates ctl/datactl53I.dat/.idx used by mvsfiles53
#2. cobfiles5A cbls cpys maps <-- creates xref/cobfiles report and also:
========================= ctl/cpyrcs1I.dat/idx & ctl/cobfilesI.dat/.idx
required by $UV/pf/IBM/mvsfiles53
See the ctl/datactl53I, cpyrcs1I,& cobfilesI files listed on page '1L6' & '1L7'.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl --> $HOME/testlibs (alias cdl='cd $HOME/testlibs)
#3. mkdir xmvsA <-- make subdir for temp files & final report
===========
#4. mvsfiles5A jcl2 <-- execute script to create report
===============
#4a. mvsfiles51 jcl2/jgl230.jcl <-- alternative script for 1 job at a time
========================== - vs all jobs in jcl2/...
#5. vi xmvsA/mvsfiles3 <-- inspect report of all files from each job
================== (in jobname sequence)
#6. vi xmvsA/mvsfiles5 <-- inspect report of all files SORTED by FILENAME
================== - more useful to determine INPUT files required
- see mvsfiles5 listed on next page after mvsfiles3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
mvsfiles3 - Add Record-sizes & Packed field indicators - 2013/12/26 10:26:28 Jobname Step Program DDname MOD Gen Rcsz pb From To * <-----DSNname------> =============================================================================== jgl100 001 CGL100 ACCTMAS S 0 0128 p GL.ACCOUNT.MASTER_ jgl100 001 CGL100 ACTLIST NCD +1 133 GL.ACCOUNT.ACNTLIST_
jgl230 001 SORT SORTIN O GL.ACCOUNT.TRAN1 jgl230 001 SORT SORTOUT NCD +1 80 002 GL.ACCOUNT.TRANS_ jgl230 002 CGL200 GLTRANS O +1 0080 001 GL.ACCOUNT.TRANS_ jgl230 002 CGL200 GLMSOLD O 0 0128 p 003 GL.ACCOUNT.MASTER_ jgl230 002 CGL200 GLMSNEW NCD +1 80 003 GL.ACCOUNT.MASTER_ jgl230 003 CGL100 ACCTMAS S 0 0128 p 002 GL.ACCOUNT.MASTER_ jgl230 003 CGL100 ACTLIST MKD +1 133 GL.ACCOUNT.ACNTLIST_
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
mvsfiles5 - Insert '*' Flags beside Esential Input files - 2013/12/26 10:26:28 Jobname Step Program DDname MOD Gen Rcsz pb From To * <-----DSNname------> ===============================================================================
jgl100 001 CGL100 ACTLIST NCD- +1 133 GL.ACCOUNT.ACNTLIST_ jgl230 003 CGL100 ACTLIST MKD- +1 133 GL.ACCOUNT.ACNTLIST_
jgl100 001 CGL100 ACCTMAS S 0 0128 p * GL.ACCOUNT.MASTER_ jgl230 002 CGL200 GLMSOLD O 0 0128 p 003 GL.ACCOUNT.MASTER_ jgl230 002 CGL200 GLMSNEW NCD- +1 80 003 GL.ACCOUNT.MASTER_ jgl230 003 CGL100 ACCTMAS S 0 0128 p 002 GL.ACCOUNT.MASTER_
jgl230 001 SORT SORTIN O * GL.ACCOUNT.TRAN1
jgl230 001 SORT SORTOUT NCD- +1 80 002 GL.ACCOUNT.TRANS_ jgl230 002 CGL200 GLTRANS O - +1 0080 001 GL.ACCOUNT.TRANS_
Note '*' files are essential Input files, only these '*' files will appear on: - next report xmvsA/mvsfiles6 summary of essential files (1 line/file) - the script xmvsA/mvsfiles7 to copy files from $CNVDATA/d2asc to $RUNLIBS/data1
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
mvsfiles6 - data file trace from MVS JCL - 2013/12/25 10:04:44 jgl100 001 CGL100 ACCTMAS S 0 0128 p * GL.ACCOUNT.MASTER_ jgl230 001 SORT SORTIN O * GL.ACCOUNT.TRAN1
# mvsfiles7 - copy essential files from $CNVDATA/d2asc/... to $RUNDATA/data1/... cp $CNVDATA/d2asc/gl.account.master_* $RUNDATA/data1 cp $CNVDATA/d2asc/gl.account.tran1 $RUNDATA/data1
#4. ksh xmvsA/mvsfiles7
===================
- execute script, copy essential files from $CNVDATA/d2asc to $RUNDATA/data1
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Before running mvsfiles5A/51, you must have run 'jcl2ksh51A' & 'cobfiles5A'.
#1. jcl2ksh51A all <-- converts all jcl0/* --> jcl1 --> jcl2 --> jcl3
============== - creates ctl/datactl53I.dat/.idx used by mvsfiles53
#2. cobfiles5A cbls cpys maps <-- creates xref/cobfiles report and also:
========================= ctl/cpyrcs1I.dat/idx & ctl/cobfilesI.dat/.idx
required by $UV/pf/IBM/mvsfiles53
ar.customer.master cpy=____________ rcs=00256 typ=RSF data=p job=jar100 prg=CAR100 ar.sales.items cpy=____________ rcs=00064 typ=LST data= job=jar200 prg=SORT gl.account.acntlist_ cpy=____________ rcs=00133 typ=LST data= job=jgl100 prg=CGL100 gl.account.master_ cpy=____________ rcs=00128 typ=RSF data=p job=jgl230 prg=CGL200 gl.account.trans_ cpy=____________ rcs=00080 typ=LST data= job=jgl230 prg=SORT gl.parms@pgl200s1 cpy=____________ rcs=_____ typ=RSF data= job=jgl200 prg=SORT py.payroll.master cpy=____________ rcs=00080 typ=RSF data=pb job=jpy200 prg=CPY200
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#2. cobfiles5A cbls cpys maps <-- creates xref/cobfiles report and also:
========================= ctl/cpyrcs1I.dat/idx & ctl/cobfilesI.dat/.idx
required by $UV/pf/IBM/mvsfiles53
unixwork1.cpy rcs=00491 pbz= bz 01=unixwork1 stline.cpy rcs=00120 pbz= 01= sdline.cpy rcs=00120 pbz= 01= saledtl.cpy rcs=00064 pbz= z 01= custmas.cpy rcs=00256 pbz=p 01=
cobfil51 ** COBOL Files Report ** Dir=cbls 2013/12/25_12:21:06 progname.cbl DDname OAM open recsz pb copybook.cpy FDname Key lines ================================================================================
car100.cbl custmas SS_ I___ 256 p custmas.cpy custmas
car100.cbl nalist L__ O___ 120 nalist
52
car200.cbl saledtl SS_ I___ 64 saledtl.cpy saledtl
car200.cbl custmas IR_ I___ 256 p custmas.cpy custmas
car200.cbl key-> cm-cust
car200.cbl salelst L__ O___ 120 sdline.cpy salelst
65
cgl100.cbl acctmas SS_ I___ 128 p acctmas
cgl100.cbl actlist L__ O___ 120 actlist
57
cgl200.cbl glmsold SS_ I___ 128 p glmsold
cgl200.cbl glmsnew SS_ O___ 128 glmsnew
cgl200.cbl gltrans SS_ I___ 80 gltrans
65
** Notes re cpyrcs1 & cobfiles1 **
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#!/bin/ksh
# mvsfiles5A - Determine essential input files for All JCLs
# - examine all data files in all JCL in subdir jcl2/...
# (after JPROCs expanded during conversion to ksh scripts)
# - by Owen Townsend, Dec10/2011, updt Sep19/2015
# mvsfiles51 - allternate job to determine essential Inputs for 1 JCL
#Sep19/2015 - create reports for All JCL in xmvsA vs mvsfiles5B & mvsfiles51
# - that create reports for each JCL in separate subdirs in xmvs/jobname
echo "Determine essential Input files needing FTP from mainframe"
jd="$1";
if [[ -d "$jd" && -d xmvsA && -d ctl ]]; then :
else echo "usage: mvsfiles5A jcl2 xmvsA"
echo " ====================="
echo " - arg1 = subdir of MVS JCL (usually jcl2/... procs expanded)"
echo " - subdir xmvsA must exist & will be cleared for reports"
echo " - subdir ctl/ must exist (will store Indexed file)"
exit 9; fi
#
echo "must have run cobfiles5A to create ctl/cobfilesI Indexed file info"
echo "must have run jcl2ksh51A or jcldata5A to create ctl/datactl53I info"
reply="x"
until [[ "$reply" == "y" || "$reply" == "n" ]]
do echo "- OK to clear all files from xmvsA/... for new reports y/n ?"
read reply
done
if [[ "$reply" == "n" ]]; then echo "job cancelled"; exit 99; fi
if [[ -d xmvsA ]]; then rm -f xmvsA/*; else mkdir xmvsA; fi
if [[ -d tmp ]]; then rm -f tmp/*; else mkdir tmp; fi
#
uvcopy mvsfiles50,fild1=$jd,filo1=xmvsA/mvsfiles0,rop=q0i7r0
#===========================================================
#
uvsort "fili1=xmvsA/mvsfiles0,rcs=128,typ=LSTt,filo1=xmvsA/mvsfiles1,key1=0(8)"
#==============================================================================
# sort by jobname only (do not change order of filenames within jobs)
#
uvsort "fili1=xmvsA/mvsfiles1,rcs=128,typ=LST,filo1=ctl/mvsfiles2I,rcs=63\
,typ=ISF,isk1=0(54d),key1=0(54),clr=0(64),mvc=0(8)0,mvc=9(44)56,mvc=55(3)9"
#==========================================================================
# sort xmvsA/mvsfiles1 by jobname+filename & load Indexed file
#
uvcopy mvsfiles53,fili1=xmvsA/mvsfiles1,filr2=ctl/mvsfiles2I,filr3=ctl/datactl53I\
,filr4=ctl/cobfilesI,filo1=xmvsA/mvsfiles3,rop=q0i7r0
#=================================================================================
#
uvsort "fili1=xmvsA/mvsfiles3,rcs=128,typ=LSTt,filo1=xmvsA/mvsfiles4\
,key1b=60(44),del1=0(1):0x20"
#====================================================================
# sort by filename, drop input blank lines,
# & insert blank lines between output filename groups
echo " " >>xmvsA/mvsfiles4 # need blank line at EOF
#
uvcopy mvsfiles54,fili1=xmvsA/mvsfiles4,filo1=xmvsA/mvsfiles5\
,filr2=ctl/datactl53I,rop=q0i7r0
#=============================================================
# copy file inserting '*' flags on required input files
# - insert recsize where absent on lines of filename groups
#
uvsort "fili1=xmvsA/mvsfiles5,rcs=128,typ=LSTt,filo1=xmvsA/mvsfiles6\
,key1u1=60(44),key2=0(8),sel1=0(1)>0x20,sel2=58(1):0x2A\
,clro=50(50),mvco=b51(50)a58"
#====================================================================
# sort by filename, unique key sort keep only 1st record of dup key set
# - drop blank lines & '*' flagged lines (required input files)
# - shift filename over beside copybook name (dropping From To steps)
#
uvcopy mvsfiles56,fili1=xmvsA/mvsfiles6,filo1=xmvsA/mvsfiles7,rop=q0i7r0
#=======================================================================
# convert final report of essential datafiles
# into script to copy from $CNVDATA/d2asc/... to $RUNDATA/data1/...
ls -l xmvsA
echo "mvsfiles5 - lists all files (with '*' on non-essential files)"
echo "mvsfiles6 - summary of essential input files"
echo "mvsfiles7 - script to copy files from \$CNVDATA/d2asc to \$RUNLIBS/data"
echo "xmvsA/... report file-names listed above, may display file-contents"
reply="x"
until [[ "$reply" == "y" || "$reply" == "n" ]]
do echo "- reply 'y' to display reports xmvsA/mvsfiles5,6,7 - y/n ?"
read reply
done
if [[ "$reply" == "y" ]]; then more xmvsA/mvsfiles[5-9]; fi
exit 0
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 2A1. | Introduction/Overview |
| 2B1. | Test Environment, RUNLIBS & RUNDATA superdirs defined in your profile |
| - allows different programmers to have their own set of Libraries & Data | |
| - best to have common set of Libraries, but different Data dirs | |
| - aliases cdl/cdd allow you to switch easily between Libs/Data dirs | |
| 2B2. | JCL converted to scripts in Part_3 & executed in Part_4 |
| - this Part_2 presents Tips & Techniques for testing & debugging |
| 2C1. | Use these tips for the Demo JCL/scripts or for your own JCL/scripts |
| See DATA file conversions in https://uvsoftware.ca/datacnv1.htm | |
| to convert your data files & FTP data from Mainframe to Unix/Linux |
| 2C2. | Converting Your files - brief review of DATAcnv1.htm. |
| Datafile Conversion Directories Required | |
| FTP data from Mainframe to Unix/Linux | |
| Re-Naming files as required |
| 2D1. | run 'testdatainit' before JCL/scripts |
| - to clear output files & make it easier to see outputs of current test | |
| - use 'joblog' to capture the log from your JCL/script | |
| - subdirs cleared by testdatainit |
| 2E1. | Test/Debug for difficult JCL/scripts |
| - save all data1/* in datasave/ & load data1/ with files for difficult job | |
| 2F2. | Iterations of test/investigate/modify as required |
| - Animator/Debugger for Micro Focus COBOL | |
| 2F3. | Check results of test/debug, investigate output files (use uvhd if packed) |
| Print-outs using 'uvlp' scripts assist test/debug |
| 2G1. | Re-Creating the GDG control file - when new JCLs added with new GDG files |
| - alternate location for gdgctl51I in $APPSADSM/ctl vs default $RUNDATA/ctl |
| 2G2. | Editing the GDG control file with no of generations desired (vs default 7) |
| - modifying text file version & reloading Indexed file used by functions | |
| in JCL/scripts |
| 2H1. | GDG files & step Restart |
| - VU JCL conversions write new GDGs into a jobtmp subdir & restore to | |
| the data subdir only if JobEnd=Normal. This allows you to rerun scripts | |
| without worrying about any new GDGs created in the failing run. | |
| - If there are steps prior to the failing step with updates in place, | |
| you can use "step Restart" at the failed step to prevent double updates. | |
| 2H2. | RERUN after failure without restart |
| 2H3. | RERUN after failure WITH RESTART |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 2I1. | jobstop & job restart |
| - activate stop at begin each step to examine files from 2nd screen |
| 2J1. | Activating 'console-logging' vs 'job-logging' for 1 JCL/script debug |
| - use console logging for long seesions running multiple jobs | |
| - captures any operator inputs (job-logging does not) |
| 2K1. | uvcmp___ - file comparison for files with packed/binary &/or no LineFeeds |
| - unix 'diff' does not work for these types of files | |
| - UV Software provides the 'uvcmp' utilities | |
| - uvcopy jobs uvcmp1,2,3 & several scripts uvcmpFA1,uvcmpFE1,etc | |
| to make the uvcopy jobs easier to run. | |
| - uvcmp prints mismatched record pairs in vertical hexadecimal | |
| flagging differences with '*'s, see sample report comparing | |
| 2 generations of gl.account.master_000001 & _000002 |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here are some tips that will help you to test/debug your JCL/scripts & COBOL programs in applications migrated from the mainframe.
Most jobs can be debugged by comparing the output files & sysout reports to the mainframe files & reports. Here are some tips for the more difficult cases. We assume that your profile has RUNDATA=$HOME/testdata (debugging in homedir), so you do not affect other programmers working. Your might also have your own set of libraries (RUNLIBS=$HOME/testlibs), OR RUNLIBS might point to a common set of libraries (JCL/scripts, COBOL programs, parms, quikjobs, etc).
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
For initial testing & self-education, you might setup testlibs, testdata,& cnvdata in your homedir. Assuming your login is userxx, here is the directory showing only the most essential subdirs. See all subdirs at JCLcnv1demo.htm#3B2.
/home/userxx <--- your homedir :-----testlibs <-- $RUNLIBS - LIBRARIEs for JCL & COBOL : :-----cbls - COBOL program source : :-----cblx - compiled COBOL programs : :-----jcls - JCL/scripts (Korn shells converted from JCLs) : : :-----testdata <-- $RUNDATA - DATA directories & subdirs : :-----data1 - datafiles : :-----joblog - programmer debug log files : :-----jobtmp - temp files for new GDGs, SYSINs, etc : :-----sysout - SYSOUT printer files (from COBOL DISPLAYs) : : :-----cnvdata <-- $CNVDATA - data CONVERSION superdir : :-----d1ebc - EBCDIC files from mainframe for conversion : :-----d2asc - files converted to ASCII : :-----cpys - COBOL copybooks : :-----maps - cobmaps (record layouts) generated from copybooks : :-----pfx1 - uvcopy jobs convert EBCDIC to ASCII (gen from cobmaps)
Note that RUNLIBS, RUNDATA,& CNVDATA are defined in your stub_profile, before calling the common_profile (in /home/appsadm/env). See profiles listed at JCLcnv1demo.htm#1C1 & 1C2. The supplied stub_profile defines these variables as follows (for initial testing, will change for production).
export RUNLIBS=$HOME/testlibs
export RUNDATA=$HOME/testdata
export CNVDATA=$HOME/cnvdata
This means that each programmer would have their own set of libraries & data, which is OK for early testing, especially if only 1 programmer on the project. For multi programmer teams, you would probably want to redefine RUNLIBS to point to a common set of libraries (not in homedirs), but you might well leave RUNDATA in $HOME/testdata, so programmers would not conflict each other.
Aliases cdl/cdd/cdc are supplied (in the common_profile), to make it easy for you to quickly switch to testlibs/testdata/cnvdata as desired.
alias cdl='cd $RUNLIBS'
alias cdd='cd $RUNDATA'
alias cdc='cd $CNVDATA'
These are even more useful after you change testlibs/testdata/cnvdata to common locations for the programmer team members - might have longer paths and you would not have to remember where.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We assume you have already converted all COBOL & JCL as documented at at JCLcnv1demo.htm#3E1 & 3F1 using super scripts that convert all copybooks & COBOL programs, and all JCLs with PROCs & PARMs.
cnvMF51Acpy all <-- convert copybooks thru all steps ============= - cpy0 --> cpy1 --> cpy2 --> cpys cnvMF51A all <-- convert COBOL programs thru all steps ============ - cbl0 --> cbl1 --> cbl2 --> cbls --> cblx
jcl2ksh51A all <-- convert ALL JCL thru all steps ============== jcl0 --> jcl1 --> jcl2 --> jcl3
--- OR for AIX, the conversion scripts would be:
cnvAIXcpyA all <-- convert copybooks for AIX COBOL =============== cnvAIXcblA all <-- convert COBOL programs for AIX COBOL ==============
jcl2ksh53A all <-- convert JCL for AIX COBOL ==============
The converted JCL/scripts are now in jcl3/..., but need to be copied to jcls/... for execution since only jcls is in $PATH (defined in the common_profile).
cp jcl3/* jcls <-- copy All scripts to execution subdir jcls/...
============== - we will do this for our test/demos
- BUT do NOT copy ALL for your "real conversion"
cp jcl3/jobnamex.ksh jcls <-- copy 1 script at a time before test/debug ========================= - recommended for your "real conversion"
You should NOT copy all jcl3/* jcls/ for your "real conversion" project. For your real conversion project, you should copy 1 script at a time when you ready to start test/debug on each script. This is an easy way to keep track of your progress (jcls/... are debugged, jcl3/... not yet)
Another reason NOT to copy all jcl3/* jcls/ when JCL/scripts re-converted:
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
jcl0 ------> jcl1 ---------> jcl2 -----------> jcl3 ------> jcls
cleanup PROC expand convert to ksh copy 1 at a time for test/debug
jcl2ksh51A all <-- script to convert ALL JCL thru all steps ============== jcl0 --> jcl1 --> jcl2 --> jcl3
You only need to rerun the 'jcl2 --> jcl3' step, using 'jclxx51' when the only changes are enhancements to JCL converter or changing options & control files.
jclxx51 jcl2 jcl3 <-- script to reconvert jcl2/* --> jcl3/... =================
To convert 1 JCL at a time, use the 'jcl2ksh51' script (vs 'jcl2ksh51A').
jcl2ksh51 jcl0/jobxx.jcl <-- convert 1 JCL from jcl0->jcl1->jcl2->jcl3->jcls ========================
'jcl2ksh51' converts thru all stages to jcl3/... & prompts for copy to jcls/... It prompts before overwriting jcls/... in case you need to save any extensive editing you have already done on some steps of the JCL/script.
JCL conversion scripts for AIX are coded with '53' vs '51' for Micro Focus.
jcl2ksh53A all <-- script to convert ALL JCL thru all steps ============== jcl0 --> jcl1 --> jcl2 --> jcl3
jclxx53 jcl2 jcl3 <-- script to convert just the JCL jcl2/* --> jcl3/...
================= - omitting cleanups & proc expansion prior to JCL convert
- save time when only changes are enhancements to JCL
converter or changing options & control files
jcl2ksh53 jcl0/jobxx.jcl <-- convert 1 JCL from jcl0->jcl1->jcl2->jcl3->jcls ========================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
It is a good idea to run the cross-references now, because they can help to resolve various problems you may encounter during testing/debugging.
xrefall cbls jcl3 ksh <-- create all cross-ref reports ===================== - in subdir xref/...
cobfiles5A cbls cpys maps <-- creates xref/cobfiles report =========================
See samples of cross-refs in JCLcnv3aids.htm#Part_1. Here are some of the most useful cross-references:
| xkshparm1 |
|
| xkshparm2 |
|
| xkshprog1 |
|
| xkshprog2 |
|
| xcobcopy1 |
|
| xcobcopy2 |
|
| xcobcall1 |
|
| xcobcall2 |
|
| cobfil51 |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
For documentation, we are using testfiles supplied with Vancouver Utilities, but for your conversion, you will need to FTP files from the mainframe & convert as illustrated with the demo files in this documentation.
See comprehensive data file conversions in 'Part_3' of this JCLcnv2real.doc. For simpler data file conversions see https://uvsoftware.ca/datacnv1.htm - especially following parts:
https://uvsoftware.ca/datacnv1.htm#Part_3
https://uvsoftware.ca/datacnv1.htm#Part_6
If you had no packed/binary fields you could FTP in TEXT MODE which automatically translates to ASCII, but the CR/LF makes records 2 bytes longer, & you either have to remove to match COBOL programs, or modify program SELECTs to 'ORGANIZATION LINE SEQUENTIAL' which allows for the CR/LF & also allows you to use unix system utilities (vi,more,lp,cat,etc) that required LineFeeds.
But most mainframes have a lot of files with packed/binary fields & it is probably best to FTP all files in BINARY MODE & use the VU conversion methods, which generate data conversion jobs automatically from the COBOL copybooks, to preserve the packed/binary fields (same for unix COBOLs as mainframe COBOLs). Note that unpacked signed fields require zoned sign conversion. Mainframe units byte positives are coded '{ABCDEFGHI' & negatives '}JKLMNOPQR'. Unix COBOL positives are coded just '0123456789' & negatives 'pqrstuvwxy'. The VU methods allow for this & an alternate version could be used with FTP text/ASCII files.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
As explained above, we are using VU test/demo files for documentation, but for your conversion, you will need to FTP files from the mainframe & convert as documented in https://uvsoftware.ca/datacnv1.htm#Part_6. Here (on pages 2C2-2C4) is a brief overview of DATA file conversion, but you must use the detailed procedures in DATAcnv1.htm.
/home/userxx <-- your homedir :-----cnvdata1 <-- cnvdata superdir for DATA conversion : : : :-----cpys <-- copybooks <--- copy from $UV/mvstest/cnvdata1/cpys/* : :-----ctl - control files : :-----d0ebc <-- EBCDIC data files <-- copy $UV/mvstest/cnvdata1/d0ebc/* : :-----d1ebc - EBCDIC data files renamed to VU conventions : :-----d2asc - ASCII data files (converted by uvcopy uvdata51) : :-----d3ebc - convert back to EBCDIC by uvdata31 : :-----d4pipe - pipe delimited to load DB tables (by genpipe1) : :-----maps - copybooks with record layouts on right side : :-----pfx1 - uvcopy jobs to convert EBCDIC to ASCII (gen by uvdata51) : :-----pfx2 - with data filenames inserted (vs copybook names) : :-----pfx3 - copied here for modify/execute : :-----pfp1 - uvcopy jobs to create pipe delimited (gen by genpipe1) : :-----pfp2 - pipe delim jobs with datafilenames inserted : :-----pfp3 - copy here before editing for multi-record-type files : :-----sqlTC1 - scripts to Create database tables (gen from copybooks) : :-----sqlTL1 - scripts to Load database tables (gen from copybooks) : :-----tmp - temp files (keep working directory clean)
cpys-------->maps---------->pfx1-------------->pfx2----------->pfx3--------->
cobmap1 uvdata51 uvdata52 copy/edit execute
convert copybooks change copybooknames
to uvcopy jobs to datafilenames
Variable-> fixed BCDIC->ASCII copy/rename
MAINFRAME -----> d0ebc ------> d1ebc ----------> d2asc ---------> $RUNDATA/data1
FTP/BINARY ASCII to d2asc pfx1/uvcopy jobs cp/renameL
rename GDG files
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here (pages 2C1-2C4) is a brief overview of how you would FTP & convert YOUR datafiles (vs demo files). You must use detailed procedures in DATAcnv1.htm.
#1. Login --> your homedir #2. cdc --> $HOME/$CNVDATA #3. cd d0ebc <-- change into the data subdir
#4. ftp xxx.xxx.xxx.xxx <-- FTP to mainframe IP#
===================
#4a. userid--> ..... <-- enter userid
#4b. passwd--> ..... <-- enter password
#4c. ascii <-- translate EBCDIC to ASCII & insert CR/LF
---OR--- (ascii is usually the default)
#4c. binary <-- ensure binary transfer
#4d. quote site RDW <-- required if any variable length files
#4e. cd ... <-- change to data files directory ??
#4f. get 'XXX.XXX.XXX' <-- get desired data files
#4g. get '...etc....' - 'single quotes' may be required
#5. Might need to change permissions on files FTP'd from mainframe
#5a. chmod 664 * <-- files should be 664
===========
//* JCL to FTP Variable-Length with RDW to Unix/Linux //ACHP01B EXEC PGM=FTP,PARM='(TRAN TRTAUGE1' //OUTPUT DD SYSOUT=* //SYSPRINT DD SYSOUT=* //INPUT DD * 30.252.99.11 21 (exit userid password cd /p2/cnv1/cnvdata/d0ebc/ pwd type i LOCSITE RDW PUT 'AAAAA.BBBBB.CCC.DDDDD.X' AAAAA.BBBBB.CCC.DDDDD QUIT /*
See more detail at https://uvsoftware.ca/datacnv2.htm#5C1 - 5C6.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
On some systems FTP'd files may be enclosed in single quotes, may be UPPER case, may have embedded blanks or '$' signs. ON unix we need to remove single quotes, remove any embedded blanks or '$' signs, & lower case is strongly recommended.
#6. mv \'XXX.XXX.XXX\' xxx.xxx.xxx <-- the hard way
================================ - see #7. the easy way
Imagine how awkward manual renaming via 'mv' commands would be if you have hundreds of files to be renamed. Vancouver Utilities has many 'rename' scripts to make mass changes to filenames easy.
#7a. renameQQ data1 <-- easy way to remove Quotes from All files in subdir
==============
#7b. renameL data1 <-- easy way to translate names to Lower case
=============
#7c. renameBU data1 <-- easy way to convert Blanks to '_' Underscores
==============
#7d. renameDU data1 <-- easy way to convert '$' signs to '_' Underscores
==============
After converting GDG files, you must rename them using the VU GDG file naming conventions - appending '_000001' for the 1st generation. For example:
#8. cp $CNVDATA/d2asc/AR.CUSTOMER.MASTER(0) $RUNDATA/data1/ar.customer.master_000001
================================================================================
- the HARD way of renaming mainframe GDG files to VU GDG standards
#8a. renameGDG d1ebc <-- script to rename all files in directory for VU GDGs
=============== (the EASY way)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
When you are testing, it is a good idea to run 'testdatainit', before you run the JCL/script being debugged. 'testdatainit' clears several subdirs, which makes it much easier to inspect the results.
Here are the testdata subdirs flagged with '0' to indicate subdirs that are cleared with testdatainit & '*' to indicate subdirs with your data files thet you would not want to clear.
/home/userxx <--- userid for testing OR your homedir :-----testdata - data files superdir : :--*--data1 - data files : :--*--datasave - backup, allow clear data1/.. to files for 1 job debug : :--*--dataMF - datafiles from MainFrame for comparison : :--*--ctl - GDG control file : :--0--jobctl - jobstop control files to debug JCL/script : :--0--joblog - programmer debug log files : :--0--jobmsgs - status msgs from JCL/scripts (step begin/end) : :--0--jobtimes - job/step times date stamped history files : :--0--jobtmp - temporary files for SYSIN data & GDG temp files : : :----JOBXX - jobtmp subdir created by jobset51 (script line 10) : : : :-----GDG : : : : :-----data1 - new GDG files restored to data1/... at EOJ : :--0--sysout - SYSOUT printer files : : :----yymmdd - date stamped subdir for print files : : : :-----JOBXX_S0010_SYSPRINT - named by job+step+DDname : :--0--tmp - tmp subdir for uvsort & misc use : :--0--tape1 - tape files become GDG files on disc
| dataMF/... |
|
| datasave/... |
|
| data1/... |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
1st run 'testdatainit' to clear output subdirs & make it easy to check results. Then run your JCL/script using 'joblog' which captures the log into the 'jblog' subdir for review if measages roll off he screen. Here is an example:
#1. Login userxx --> $HOME (/home/userxx)
============
#2. cdd --> $HOME/testdata
======================
#3. testdatainit <-- clear subdirs of prior run data
============ - also prompts to reload the GDG control file
#4. joblog jobxx.ksh <-- use joblog to run script to be tested
================ - saves console messages in joblog/...
The subdirs cleared are: jobctl, joblog, jobmsgs, jobtmp, sysout,& tmp. We will discuss in order of importance:
| Note |
|
| Note |
|
llr sysout <-- will show only new files from your latest run
- because testdatainit has cleared outputs of prior runs
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
jobxx.ksh start=S0030 stop=S0030
================================
- would run only step 3
- omit the stop=... to run all remaining steps
- would use to restart production jobs after failing step
| Note |
|
#7. llr jobtmp <-- list all filenames in jobtmp/ such as GDG files
==========
#8. l jobtmp/JOBXX/GDG/data1/ <-- list any GDG files left in jobtmp
========================
- JobEnd=Normal moves new GDG files back to data1/... subdir
- JobEnd=AbTerm leaves new GDG files in jobtmp/JOBXX/GDG/data1/...
so you can fix problem & rerun without removing new GDGs from data1/...
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We assume that you have been test/debugging many jobs & then find a difficult job that requires these exceptional procedures. The data1/... subdir could have hundreds of files from prior tests; we can save them to datasave/..., clear all data1/* & copyback only the files required for the difficult job being debugged.
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdd --> $HOME/testdata
======================
#3. mkdir datajobxx <-- mkdir for files required in jobxx
=============== (problem job to be debugged)
#3a. cp data1/fileA datajobxx/ <-- select files required for jobxx
=========================
#3b. cp data1/fileB datajobxx/ ... etc ...
=========================
#4. mkdir datasave <-- make backup subdir for existing datafiles
============== (if not already existing)
#5. cp data1/* datasave <-- backup all existing data1/* files
===================
#7a. cp datasave/* data1
===================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
This assumes that you have completed the preparations suggested on the previous pages, especially the backup of data1/* files to datasave/..., clear all data1/*, & copyback only the files required for the difficult job being debugged. You could now repeat the following steps for each cycle of test/debug that might be required for a difficult job.
#1. cdd --> ensure we are in $RUNDATA
#2. rm -f data1/* <-- remove all files from data1/...
=============
#3. cp datajobxx/* data1 <-- store data1/ files for jobxx only
====================
#4. testdatainit <-- clear files from all temporary subdirs
============ (jobctl,joblog,jobmsgs,jobtmp,sysout,tmp,wrk)
#5. export ANIM="+A" <-- activate Micro Focus COBOL animator (debugger)
================
#6. joblog jobxx.ksh <-- test/debug jobxx
================
#7. reply to animation prompts
==========================
The Micro Focus Animator is a marvelous debugging tool for COBOL programs. It stops at 1st instruciton & displays a menu of comamnds. Here are a few:
--> s <-- Step (execute current high-lighted statement)
- for Perform: go into paragraph or section
--> ps <-- Perform Statement
- do not animate paragraph/section (appears to step over it)
--> qc <-- Query the fieldname under the Cursor
- move cursor (with arrow keys) to the desired fieldname
--> lc <-- Locate definition of fieldname under the Cursor
--> g <-- Go, steps continuously (but slowly) thru instructions
--> 9 <-- (while Go is progressing), enter digit 2-9 to speed up
--> z <-- Zoom thru remainder of program (without animating)
--> Esc <-- Escape to exit program
--> y <-- requires 'y' response to actually exit
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#8. vi joblog/jobxx.log <-- view the joblog
===================
#9a. vi data1/... <-- check data files with 'vi'
============ - IF LineFeeds present (no Packed fields)
#9b. uvhd data1/... <-- use 'uvhd' if no LF's
============== - &/or if Packed/Binary fields present
#10a. llr jobtmp <-- list jobtmp files created
========== - intermediate files & new GDG's
(new GDGs NOT restored to data1/ if Abterm)
#10b. vi jobtmp/JOBXX/GDG/... <-- investigate GDG files
=======================
#11a. llr sysout <-- list sysout files created
========== - from COBOL 'DISPLAY's
#11b. vi sysout/JOBXX/... <-- investigate sysout files as desired
===================
#11c. uvlp18 sysout/JOBXX/... <-- print sysout files as desired
======================= - uvlp18 prints 18 cpi (132 cols on 8" wide)
#11d. uvlp14L sysout/JOBXX/... - use uvlp12L (Landscape) 14cpi 150 cols 11"
========================
#12. cdl --> $RUNLIBS <-- switch to $RUNLIBS
#13a. vi jcls/jobxx.ksh <-- modify the JCL/script if required
=================
#14a. vi cbls/cblxx.ksh <-- modify COBOL programs if required
=================
#14b. mfcbl1 cblxx.cbl <-- recompile the COBOL program
================
-------------> repeat steps 1 - 14 as required <---------------
#15. cp datasave/* data1 <-- restore all data1/* files when jobxx debugged
===================
#16. export ANIM="" <-- de-activate Micro Focus COBOL animator
==============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Hard-copy listings can help you debug difficult jobs & document the problems for other team members. Vancouver Utilities provides several scripts (based on the uvlist utility documented at uvlist.htm. These scripts are great for JCL/script & program listings since the page headings capture vital information (filename, date, userid, page#, line# of 1st line on page,etc). Here are a few of the most useful scripts:
| uvlp12 |
|
| uvlp12D |
|
| uvlp14 |
|
| uvlp14L |
|
| etc |
|
#1. uvlp14 joblog/jobxx.log <-- print the joblog for better study
======================= - see 'uvlp' scripts in uvlist.htm
#2. lslp data1 <-- print data1/... to study with jobxx.log
========== - 'lslp' script does ls -l to tmp/ & uvlp12 tmp file
#3. llr sysout <-- display printer filenames in sysout/yyyymmdd/...
==========
#4. uvlp14 sysout/yyyymmdd/JOBXX_S0010_DDname
=========================================
- print sysout files
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
If you have not yet edited gdgctl51 & reloaded gdgctl51I, then you could re-create gdgctl51 from all JCL by running 'uvcopy jclgdgctl51' separately from script 'jcl2ksh51A'. You might do this if you had added a few JCLs via jcl2ksh51 (vs jcl2ksh51A).
#1. Login userxx --> /home/userxx #2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3. uvcopy jclgdgctl51,fild1=jcl3,filo1=ctl/gdgctl51
================================================
- extract all GDG file-names from converted JCL/scripts
| NOTE |
|
#4. cp ctl/gdgctl51 $GDGCTL/ctl/ <-- copy to GDG control directory
============================ - GDGCTL $RUNDATA or $APPSADM
#5. uvcopy gdgload51,fili1=$GDGCTL/gdgctl51,filo1=$GDGCTL/gdgctl51I
===============================================================
#5a. uvcopy gdgload51 <-- same but easier (files default as shown above)
================
If you reconvert all JCL (using script 'jcl2ksh51A'), you will NOT overwrite the GDG control file & lose any manual edits to modify the number of generations, because the script tests for existing $GDGCTL/gdgctl51 & will not overwrite.
If you only needed to convert a few additional JCL's you could use script 'jcl2ksh51' which converts 1 at a time & does not recreate the GDG control file. You could then edit the existing gdgctl51 adding any new GDG files & reload the Indexed file by rerunning 'uvcopy gdgload51'.
The location of the GDG control file is determined by the common_profile (stored at /home/appsadm/env/common_profile & called by .profile or .bash_profile). Here are lines 61-62 of $APPSADM/env/common_profile.
export GDGCTL=$RUNDATA/ctl #<-- default location # export GDGCTL=$APPSADM/ctl #<-- could change to this ?
If you have multiple RUNDATA directories you might want to activate 'export GDGCTL=$APPSADM/ctl' & #comment out 'export GDGCTL=$RUNDATA/ctl' so you would have only 1 GDG control file vs multiple.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The GDG control file is 1st created as part of the 'jcl2ksh51A' script that converts all JCL to Korn shell scripts. GDG filenames are identified by suffixes (0),(+1),etc & are extracted from all JCL, sorted dropping duplicates,& loaded into an Indexed file for lookup when the JCL/scripts are executed - to determine latest & next generation,& also to determine how many generations to maintain. These lookups & determinations are performed by functions 'exportgen0' & 'exportgen1'.
Here is the GDG file created from the demo JCL/scripts listed in Part_2. This is the text file that can be edited to modify the number of generations before loading into the Indexed file 'gdgctl51I' (that is looked up by the exportgen0/exportgen1 functions in the JCL/scripts).
gl.account.acntlist_ gdg=07 gl.account.master_ gdg=07 gl.account.trans2_ gdg=07 gl.account.trans_ gdg=07
As shown above the number of generations defaults to 7, You ca edit the file to set the generations as desired for each file. Note that the location of the GDG control file defaults to $RUNDATA/ctl/gdgctl51 but is configurable via "export GDGCTL=..." (discussed later).
#1. Login userxx --> /home/userxx
#2. cdd --> /home/userxx/testdata (alias='cd $RUNDATA)
===
#3. vi ctl/gdgctl51 <-- edit to modify no of generations
===============
#4. uvcopy gdgload51
================
- reload Indexed file for use by exportgen1, exportgen0,& jobend51
I/O files are defaulted within the 'gdgload51' uvcopy job as follows:
#4a. uvcopy gdgload51,fili1=$GDGCTL/gdgctl51,filo1=$GDGCTL/gdgctl51I
===============================================================
Since the common_profile defines "export GDGCTL=$RUNDATA/ctl", the result is:
#4b. uvcopy gdgload51,fili1=$RUNDATA/ctl/gdgctl51,filo1=$RUNDATA/ctl/gdgctl51I
=========================================================================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The VU JCL conversion writes new GDGs into $RUNDATA/jobtmp/$JOBID2/GDG/data1/... and restores the new GDGs to $RUNDATA/data1/... at S9000 JobEnd=Normal. If any step fails, the job ends at S9900 JobEnd=AbTerm & any new GDG files in jobtmp/... are not restored to data1/...
The logic above means that you can simply rerun JCL/scripts with GDG files and not worry about any new GDGs created in the failing run. However you do need to know if there are any non-GDG files that are updated in place (vs updated by copying to a newfile). If there are updates in place prior to the failing step, then you can use "step Restart" to prevent double updates. This assumes it was not the failing step doing an update in place, in which case you would need to restore that file from a backup made before the job started.
We will demo new GDG file creation & step restart using JCL/script 'jgl200.ksh' (listed at JCLcnv1demo.htm#2D1 & see normal joblog at JCLcnv1demo.htm#4F4 Step1 sorts a transaction file to a GDG file which updates a GDG master file on the step2 COBOL program. We can cause 'AbTerm' by moving the COBOL executable from $RUNLIBS/cblx to tmp/... & return after failed run to demo recovery.
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl --> $HOME/testlibs (alias cdl='cd $HOME/testlibs)
#3. mv cblx/cgl200.int tmp <-- move step2 program out to tmp/...
====================== - to cause failure on step2
#4. cdd --> $HOME/testdata
#5. l data1 <-- list existing files in $RUNDATA/testdata/data1
======= - will show only files for jgl200.ksh
- only 1 generation of each file existing
-rw-rw-r-- 1 userxx apps 13952 Oct 18 15:11 gl.account.master_000001 -rw-rw-r-- 1 userxx apps 1600 Oct 18 15:11 gl.account.trans_000001
#6. joblog jgl200.ksh <-- execute the job (step2 will fail)
================= - joblog listed below (just step2 & AbTerm msgs)
******** Begin Step S0020 cgl200 (#2) ******** gen+1 GLTRANS=jobtmp/JGL200/GDG/data1/gl.account.trans_000002 gens=07 gen0 GLMSOLD=data1/gl.account.master_000001 insize=4.0K gen+1 GLMSNEW=jobtmp/JGL200/GDG/data1/gl.account.master_000002 gens=07 Executing--> cobrun -F /home/userxx/testlibs/cblx/cgl200 Load error : file '/home/userxx/testlibs/cblx/cgl200' error code: 173, pc=0, call=1, seg=0 173 Called program file not found in drive/directory ERR: step#S0020 cgl200 abterm 255 JobEnd=AbTerm, JCC=255,StepsX/L=2/S0020 GDG files NOT moved from jobtmp/subdirs to /home/userxx/testdata/subdirs
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#7. cdl --> $HOME/testlibs
#8. mv tmp/cgl200.int cblx <-- move step2 program back to execution library
======================
#9. cdd --> $HOME/testdata
#10. joblog jgl200.ksh <-- re-execute the job, will get messages
================= about GDG files in jobtmp from prior failure
WARN: files in jobtmp/GDG subdirs (from prior AbTerm ERR?)
---- files in jobtmp/JGL200/GDG/*/* listed below:
jobtmp/JGL200/GDG/data1/gl.account.trans_000002
If NO restart by step#, GDG files in jobtmp/... will be cleared
- allows rerun from begin job with no worry about GDGs
If RESTARTing by step#, example--> jobname.ksh start=S0050
- GDG files in jobtmp/... will NOT be cleared
- will be available to steps after restart step#
- will be restored to data1/... subdir at JobEnd=Normal
--> <-- null entry to continue
- - - lines removed to JobEnd=Normal - - -
JobEnd=Normal, StepsExecuted=1, LastStep=S0020
#11. l data1 <-- list existing files in $RUNDATA/testdata/data1
======= - will show only files for jgl200.ksh
- now 2 generations each
-rw-rw-r-- 1 userxx apps 13952 Oct 18 15:11 gl.account.master_000001 -rw-rw-r-- 1 userxx apps 13952 Nov 26 12:15 gl.account.master_000002 -rw-rw-r-- 1 userxx apps 1600 Oct 18 15:11 gl.account.trans_000001 -rw-rw-r-- 1 userxx apps 1600 Nov 26 12:15 gl.account.trans_000002
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#10. joblog jgl200.ksh start=S0020 <-- re-execute job, with RESTART at step2
==================***********
- will get same msgs re GDG files
- BUT, notice the 'WARN: **START**' msg now appears
WARN: files in jobtmp/GDG subdirs (from prior AbTerm ERR?)
---- files in jobtmp/JGL200/GDG/*/* listed below:
jobtmp/JGL200/GDG/data1/gl.account.trans_000002
If NO restart by step#, GDG files in jobtmp/... will be cleared
- allows rerun from begin job with no worry about GDGs
If RESTARTing by step#, example--> jobname.ksh start=S0050
- GDG files in jobtmp/... will NOT be cleared
- will be available to steps after restart step#
- will be restored to data1/... subdir at JobEnd=Normal
--> <-- null entry to continue
WARN: **START** at start=S0020
- - - lines removed to JobEnd=Normal - - -
JobEnd=Normal, StepsExecuted=1, LastStep=S0020
#11. l data1 <-- list existing files in $RUNDATA/testdata/data1
======= - will show only files for jgl200.ksh
- still 2 generations each
-rw-rw-r-- 1 userxx apps 13952 Oct 18 15:11 gl.account.master_000001 -rw-rw-r-- 1 userxx apps 13952 Nov 26 12:15 gl.account.master_000002 -rw-rw-r-- 1 userxx apps 1600 Oct 18 15:11 gl.account.trans_000001 -rw-rw-r-- 1 userxx apps 1600 Nov 26 12:15 gl.account.trans_000002
See complete documentation re GDG files at JCLcnv4gdg.htm.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The JCL converter inserts a call to function 'stepctl51' st the begining of each step. This function normally does nothing, but can be activated by 'jobstop' to stop at the begining of each step (or a specified step) & wait for the operator to enter 'go' or 'clear'.
stepctl51 # test oprtr jobstop/jobclear #======== # inserted at begining of each step by JCL converter option 'b2'
Pausing at the begining of each step allows you to investigate the files in data1/... & jobtmp/JOBXX/GDG/data1/... that might be deleted at the end of the step. Also see page '5B1' which suggested you might use the JCL converter option to #comment out 'rm's (converted from original 'DELETE's) & then later uncomment with mass change scripts.
#1. Login --> /home/userxx #2. cdd --> $HOME/testdata
---> instructions omitted here (see page '2H2')
#6a. jobstop jobxx.ksh <-- activate stop at begin each step
================= - stores control record in jobctl/jobxx.ctl
#6b. joblog jobxx.ksh <-- execute the job
================ - will get following display at begin each step
--> jobxx.ksh paused by job control file: jobctl/jobxx.ctl
- job control record: jobxx.ksh S0000 111125_124512
- waiting until reply 'go' or 'clear'
#7. Login on a 2nd screen --> $HOME #8. cdd --> $HOME/testdata
#9a. l data1 <-- investigate data files on screen#2
======= - while jobxx paused on screen#1
#9b. l jobtmp/JOBXX/GDG/data1/ <-- investigate GDG files in jobtmp
=========================
The 'jobstop' script may also specify the 1st step# to stop (art end of step)
#6a. jobstop jobxx.ksh stop=S0030 <-- store jobctl file with stop step#
============================
#6b. joblog jobxx.ksh <-- execute the job
================ - will stop at end of step S0030
& subsequent steps if you reply 'cont'inue
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
As of Feb 2012, the 'job restart' facility has been enhanced to include the 'job stop' facility. This means you no longer have to precede your job execute (with or without joblog) by the 'jobstop' script to store the jobctl/file. You can restart a job at a specified step# & stop it at the END of a step#, Here are some examples:
#1. jgl310.ksh <-- runs all 6 steps of this demo job
==========
#1a. joblog jgl310.ksh <-- same with joblogging
================= - will omit 'joblog' from following examples
- BUT you should use joblog for test/debug
#2. jgl310.ksh start=S0030 <-- rerun job starting at step 3
====================== - will run to end of job with no stops
#3. jgl310.ksh start=S0030 stop=S0040
=================================
- rerun job starting at step 3 & stopping at END of step 4
- displays following choices:
120210:155447:JGL310: jgl310.ksh paused by job control file: jobctl/jgl310.ctl 120210:155447:JGL310: - job control record: jgl310.ksh S0030 120210_155447 120210:155447:JGL310: - waiting for reply: cont(inue), clear, endok, endab
cont - continue (execute) to next step, stop again,& reprompt as above clear - clear the jobctl/jgl310.ctl file, will execute to end job with no stops endok - goto Normal end of job, will restore any new GDG's endab - goto AbNormal end of job, will NOT restore any new GDG's
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
If desired you could activate console-logging & rerun the demo jobs.
Console-Logging will capture the entire login session from login to logout vs the 'joblog' script intended for programmer test/debug to capture the log for 1 job at a time.
Note that console-logging captures all screen I/O including any reponses to prompts & commands the operator might do between running jobs.
See complete details of console logging at ADMjobs.htm#Part_6 but here is the short version of activating. ADMjobs.htm#6D1 describes the subdirs required in /home/appsadm where logs are collected. We will assume that you are 'userxx'.
#1. login --> your homedir
#2a. mkdir $APPSADM/log1/userxx <-- subdir current log being created #2b. mkdir $APPSADM/log2/userxx <-- subdir current months processed logs #2c. mkdir $APPSADM/log3/userxx <-- subdir last months logs
#3. vi .profile <-- edit your profile
===========
#3a. uncomment the 9 '##' lines near the end of your profile
- see profile listing on page JCLcnv1demo.htm#1C1
#3b. :wq
#4. Logout & Log back in to start console logging
#5. Run some jobs (jar100.ksh for example) #5a. cdd - change to $RUNDATA directory #5b. l - list subdirs in $RUNDATA #5c. jar100.ksh - run 1st job (COBOL program car100.cbl) #5d. l data1 - list data subdir
#6. Logout/Login to process the log
- copies from $APPSADM/log1/userxx/date_time to $APPSADM/log2/userxx/.
removing screen control codes that would make log unreadable
#7. logview <-- script lists your available log filenames
======= - prompts for file# to view (#1 is latest)
#7a. --> 1 <-- enter '1' to see latest log
#7b. --> 0 <-- enter '0' to quit
#8. l $APPSADM/log2/userxx/ <-- can list your log files directly
======================= vs logview script
#8a. vi $APPSADM/log2/userxx/yymmdd_HHMMSS <-- inspect latest log file
=====================================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'job-logging' is good for test/debug (captures log for 1 job at a time), but 'console-logging' is better for production, because it captures everything, inputs as well as outputs, what commands the operator did between jobs. Here is a sample - I have numbered the operator's commands on right side:
#1. cdd <-- change to $RUNDATA #2. l <-- list files in $RUNDATA #3. l data1 <-- list files in data1/ (before running job) #4. jar100.ksh <-- run JCL/script #5. l data1 <-- list files after running job
<@uvsoft4:userxx:/home/userxx> cdd <-- #1. <@uvsoft4:userxx:/home/userxx/testdata> l <-- #2. drwxrwxr-x 2 userxx apps 4096 Oct 17 10:49 ctl drwxrwxr-x 2 userxx apps 4096 Oct 17 09:48 data1 <-- data files subdir drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 joblog drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 jobmsgs drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 jobtmp drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 sysout drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 tmp drwxrwxr-x 2 userxx apps 4096 Oct 17 09:27 wrk
<@uvsoft4:userxx:/home/userxx/testdata> l data1 <-- #3. -rw-rw-r-- 1 userxx apps 8192 Oct 17 09:48 ar.customer.master -rw-rw-r-- 1 userxx apps 1280 Oct 17 09:48 ar.sales.items -rw-rw-r-- 1 userxx apps 13952 Oct 17 09:48 gl.account.master_000001 -rw-rw-r-- 1 userxx apps 1600 Oct 17 09:48 gl.account.trans_000001
<@uvsoft4:userxx:/home/userxx/testdata> jar100.ksh <-- #4.
==========
111017:105233:JAR100: Begin Job=JAR100
111017:105233:JAR100: /home/userxx/testlibs/jcls/jar100.ksh
111017:105233:JAR100: RUNLIBS=/home/userxx/testlibs
111017:105233:JAR100: RUNDATA=/home/userxx/testdata
111017:105233:JAR100: JTMP=jobtmp/JAR100 SYOT=sysout/JAR100
111017:105233:JAR100: RUNDATE=20111017
111017:105233:JAR100: ******** Begin Step S0010 car100 (#1) ********
111017:105233:JAR100: file: CUSTMAS=data1/ar.customer.master fsize=8.0K
111017:105233:JAR100: file: NALIST=data1/ar.customer.nameadrs.list100 fsize=
111017:105233:JAR100: file: SYSOUT=sysout/JAR100/S0010_SYSOUT fsize=
111017:105233:JAR100: Executing--> cobrun -F /home/userxx/testlibs/cblx/car100
111017:105233:JAR100: Job Times: Begun=10:52:33 End=10:52:33Elapsed=00:00:00
111017:105233:JAR100: EOF filr01 rds=3 size=6144: /home/userxx/testdata/ctl/gdgctl51I
111017:105233:JAR100: JobEnd=Normal, StepsExecuted=1, LastStep=S0010
<@uvsoft4:userxx:/home/userxx/testdata> l data1 <-- #5. -rw-rw-r-- 1 userxx apps 8192 Oct 17 09:48 ar.customer.master -rw-rw-r-- 1 userxx apps 2858 Oct 17 10:52 ar.customer.nameadrs.list100 -rw-rw-r-- 1 userxx apps 1280 Oct 17 09:48 ar.sales.items -rw-rw-r-- 1 userxx apps 13952 Oct 17 09:48 gl.account.master_000001 -rw-rw-r-- 1 userxx apps 1600 Oct 17 09:48 gl.account.trans_000001
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'diff' is a marvelous utility, but does not work for files without lineFeeds & with packed/binary fields present. For these files UV Software provides the 'uvcmp' utilities - uvcopy jobs uvcmp1,2,3 & several scripts uvcmpFA1, uvcmpFE1, etc to make the uvcopy jobs easier to run.
Here is an example using JCL/script 'jgl200', listed at JCLcnv1demo.htm#2D1 & executed on page 4F4). jgl200 updates the account.master file with a transaction file while copying to a new output file. Since the account.master file is a GDG file, it is easy to compare the old & new master files without having to save the input file.
#1. Login mvstest1 --> /home/mvstest1 #1. or Login yourself (userxx) --> your homedir (/home/userxx)
#2. cdd --> $HOME/testdata (alias cdd='cd $HOME/testdata)
#3. joblog jgl200.ksh <-- execute account.master update
================= - see joblog at JCLcnv1demo.htm#4F4
#4. l data1 <-- list data1/... subdir after execution
=======
-rw-rw-r-- 1 userxx apps 13952 Oct 18 15:11 gl.account.master_000001
-rw-rw-r-- 1 userxx apps 13952 Dec 13 11:31 gl.account.master_000002
-rw-rw-r-- 1 userxx apps 1600 Oct 18 15:11 gl.account.tran1
-rw-rw-r-- 1 userxx apps 1600 Oct 18 15:11 gl.account.trans_000001
#5. mkdir rptcmp <-- make subdir for comparison report
============ (if not already present)
#6. uvcmpFA1 data1/gl.account.master_000001 data1/gl.account.master_000002 r128
===========================================================================
- compare the before & after update files, 'r128' specifies the record-size
- may omit option 'r' on the command line & specify when prompted
- writes difference report to rptcmp/gl.account.master_000001
(ie - named same as 1st file compared)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#7. vi rptcmp/gl.account.master_000001 <-- inspect difference report
================================== - 1st 2 record pairs listed below:
uvcmp1 - compare 2 files, print mismatched records, '*' flag diffs
2011/12/13_11:32:34 uop=q1p30r256s6t500000u3x2y0q1r128
recsize reccount file-size typ Report=rptcmp/gl.account.master_000001
1: 128 109 13,952 RSF File1=data1/gl.account.master_000001
2: 128 109 13,952 RSF File2=data1/gl.account.master_000002
1 2 3 4 5 6
f#record#byte# 0123456789012345678901234567890123456789012345678901234567890123
===============================================================================
1 1 0 11100 11100Royal Bank Lynn Valley ....b.20090131
2333332222222333335676624666247662566667222222220010683333333322
0111000000000111002F91C021EB0C9EE061CC590000000000972C2009013100
******
2 1 11100 11100Royal Bank Lynn Valley ....b.20111213
2333332222222333335676624666247662566667222222220010683333333322
0111000000000111002F91C021EB0C9EE061CC590000000000972C2011121300
******
- - - 5 record pairs omitted - - -
==================== EOF or StopPrint/StopRead count reached ============== F1Count=109, F2Count=109, StopPrint=6, StopRead=500000 F1Reads=109, MisMatches=109, MisMatsPrinted=6, Recsize=128
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 3A1. | Control-Files for JCL conversion to supply data-file-info |
| - that may be missing in JCL (record-sizes,file-types,etc) | |
| - required for script conversions of SORTs,IDCAMs,IEBGNENERs,etc |
| 3B1. | 1st conversion of COBOL & JCL (before updating control files) |
| - operating instrucitons (review, 1st presented at '3D1', 3E1, 3F1) |
| 3B2. | control-files before 1st conversion |
| - summary listing |
| 3B3. | capturing COBOL info for JCL conversion (record-sizes, file-types) |
| 3B4. | control files created by 1st conversion (in subdir ctl/...) |
| - datajcl51, datajcl52 extracted from the JCL |
| 3B5. | JCL converter script 'jcl2ksh51A' does everything |
| - calls 'jcldata51A' to create datajcl51 & datajcl52 | |
| & load into Indexed file datactl53I.dat/.idx for JCL converter | |
| - calls 'jclxx51' for JCL conversion to Korn shell scripts | |
| - updates datactl53I with datafile info from COBOL conversions |
| 3B6. | datactl53I control file before & after update by 1st JCL conversion |
| 3C1. | using LISTCAT mainframe reports to improve JCL conversion |
| - transfer, translate, store multiple reports in subdir cat0/... |
| 3C2. | LISTCAT report sample (mock-up for VU demo files) |
| 3C3. | control files extracted from LISTCAT reports |
| - processed to subdirs cat1/..., cat2/... | |
| - summarized to 1 report in ctl/datacat51 | |
| - loaded into Indexed file ctl/datacat52I.dat/.idx | |
| - merged into ctl/datactl53I.dat/.idx by jcldata51A |
| 3D1. | Problem - record sizes missing from LISTCAT report |
| Solution - an extra LRECL report from mainframe to be converted | |
| to keyword format & combined with other control files |
| 3E1. | manual coded control file for missing record-sizes,file-types,etc |
| 3E2. | converting Excel spreadsheets of datafile info to control files |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 3F1. | Re-Converting JCL with LISTCAT, datamisc,& dataxls |
| - using 'alldiff2' script to see differences made by re-conversion | |
| - confirm chages as expected & no unintended changes |
| 3F2. | dropping unwanted lines on alldiff2 report with vi & diffdrop2 |
| 3F3. | - alldiff2 report before & after vi & diffdrop2 |
| 3F4. | re-convert alternative to save time if hundreds of JCLs,PROCs,PARMs |
| - use jcldata51A & jclxx51 (vs do everything jcl2ksh51A) |
| 3F5. | JCL converter control file ctl/datactl53I |
| - before & after 2nd conversion |
| 3G1. | Modifying JCL converter 'options' |
| - example to change fixed length data files to Line Sequential | |
| - if you have very few files with packed/binary fields |
| 3H1. | Modifying JCL converter 'data info control files' & reconverting |
| - example to correct or supply missing record-sizes & file-types |
| 3I1. | JCL converter control file summary |
| 3J1. | setup alternate sets of conversion directories |
| - selected sets of JCL/COBOL for testing, education, etc |
| 3K1. | Determine Essential INPUT files for any 1 JCL/script |
| script 'mvsfiles51' selects essential input files | |
| - ignoring output files & intermediate created/used within JCL |
| 3L1. | Inserting commands to print files |
| - using uvlist & 'uvlp' scripts to insert PCL5 for laser printers | |
| - uvlp13LD to get mainframe 132 cols 66 lines on 8 1/2 x 11 | |
| (Landscape & Duplex) |
JCLcnv2real.htm#4J1 JCL converter 'options' control file (ctl/jclunixop51)
- copied from $UV/ctl/... to $RUNLIBS/ctl/... by 'copymvsctls'
after initial setup of conversion directories by 'mvslibsdirs'
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here are the more relevant subdirs we will use to illustrate customizing JCL conversions by modifying control files & re-converting. See the jclunixop51 JCL converter options control file at JCLcnv2real.htm#4J1.
/home/uvadm/mvstest/testlibs <-- test/demo libraries supplied in /home/uvadm/... /home/userxx <-- can copy to your homedir :-----testlibs : :------cat0 - LISTCAT reports from mainframe (see '3C1') : :------cat1 - LISTCAT reports converted to unix control files : :------cat2 - LISTCAT reports translated to lower case : :------cbl0 - mainframe COBOL programs : :------cbl1/2 - intermediate subdirs used in conversion : :------cbls - converted COBOL programs : :------cblx - compiled COBOL programs : :------cpy0 - mainframe COBOL copybooks : :------cpy1/2 - intermediate subdirs used in conversion : :------cpys - copy here (standard copybook library) : :------ctl - conversion control files : : :---add/... <-- subdir for ADDITIONAL data control files : : :-----add/datactl53a <-- saved by JCL converter : : :-----add/datactl53b <-- may copy a to b & update file types : : : rec-sizes, etc for subsequent conversions : : :-----add/datamisc51 <-- add info missing on 1st convert : : :-----add/dataxls51 <-- add info from Excel spreadsheet : : :---datacat51 <-- DATA file info from LISTCAT files : : :---datactl53I <-- all DATA file info merged to Indexed file : : :---gdgctl51 <-- GDG datafile generations to maintain : : :---jclunixop51 <-- control file/options for JCL converter : :------jcl0 - mainframe JCL : :------jcl1 - intermediate conversion 73-80 cleared : :------jcl2 - PROCs expanded from procs : :------jcl3 - JCLs converted to Korn shell scripts : :------jcls - copy here manually 1 by 1 during test/debug : :------parm0 - control cards & includes (SORT FIELDS, etc) : :------parms - control cards with 73-80 cleared : :------proc0 - test/demo PROCs supplied : :------procs - will be merged with jcl1, output to jcl2 : :------tmp - tmp subdir used by uvsort & various conversions : :------xref - cross-references
You can modify various options in ctl/jclunixop51 & re-convert. The complete conversion of all JCL components can be pictured as follows:
jcl0 ------> jcl1 ---------> jcl2 -----------> jcl3 ------> jcls
cleanup PROC expand convert to ksh copy 1 at a time for test/debug
jcl2ksh51A all <-- script to convert ALL JCL thru all steps ============== jcl0 --> jcl1 --> jcl2 --> jcl3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here are the control files after 'copymvsctls' (see '3D1') has copied initial control files from $UV/ctl/... before the 1st COBOL & JCL conversion. See copymvsctls listed at JCLcnv2real.htm#4C0
/home/userxx/testlibs/ :-----ctl/ : :----add/ : : :-----dummy_readme : : :-----add/datamisc51 <-- add info missing on 1st convert : : :-----add/dataxls51 <-- add info from Excel spreadsheet : :-----cnvcob5.tbl <-- COBOL conversion search/replace table : :-----cobdirectives <-- Micro Focus COBOL compile Directives : 0 :-----cobfil55bI.dat <-- COBOL datafile info (initially empty) : :-----cobfil55bI.idx : 0 :-----datacat52I.dat <-- LISTCAT datafile info (initially empty) : :-----datacat52I.idx : :-----datamisc51 <-- you can edit with misisng record sizes : :-----dataxls51.psv <-- file of record sizes from Excel spread sheet : :-----extfh.cfg <-- Micro Focus COBOL Extended File Handler config : :-----jclunixop51 <-- JCL converter control-file : :-----utilities
Here are the control files after the 1st conversion - with comments on right only for files changed by 1st conversion.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
/home/userxx/testlibs/ :-----ctl/ : :-----add/ : : :-----datactl53a <-- copied/renamed from ../datactl53I.dat : : :-----datactl53b <-- can copy/rename/modify recsizes & filetypes : : :-----dummy_readme : :-----cnvcob5.tbl : :-----cobdirectives : :-----cobfil55a <-- COBOL datafile info stored by cnvMF51A : :-----cobfil55aI.dat - to Identify ORG Line Seqntl (print) files : :-----cobfil55aI.idx : :-----cobfil55b <-- COBOL datafile info stored by cnvMF51A : :-----cobfil55bI.dat - to supply record-sizes to JCL converter : :-----cobfil55bI.idx : :-----dataadd51 <-- datafile info from all ctlfiles in add/... : :-----dataadd52I.dat - added to datactl53I used by JCL converter : :-----dataadd52I.idx : :-----datacat51 <-- datafile info from LISTCAT cat2/... : :-----datacat52I.dat - added to datactl53I used by JCL converter : :-----datacat52I.idx : :-----datactl53 <-- datafile info for JCL converter (sequential) : :-----datactl53I.dat - Indexed file version used by JCL converter : :-----datactl53I.idx : :-----datamisc51 <-- you can edit with misisng record sizes : :-----dataxls51.psv <-- file of record sizes from Excel spread sheet : :-----datajcl51 <-- datafile info extracted from all jcl2/... : :-----datajcl52 - 1st input to create datactl53 for JCL converter : :-----extfh.cfg : :-----gdgctl51 <-- initial GDG control file created by jcl2ksh51A : : - copy to $RUNDATA/ctl, modify no of generations : : & load Indexed file for JCL/script executions : :-----jclunixop51 : :-----utilities
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We have already performed the 1st conversion as documented in JCLcnv1demo.htm Here is a brief review of the setup & conversion scripts (showing scripts for Micro Focus COBOL & AIX equivalents)
#1. login yourself --> /home/userxx
#2. mkdir testlibs <-- make superdir for JCL/COBOL subdirs
#3. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#4. mvslibsdirs <-- setup 30 subdirs for JCL & COBOL conversions
===========
#5. copymvsctls <-- script to copy control files from /home/uvadm/ctl
=========== to $RUNLIBS/ctl/...
- see copymvsctls listed at JCLcnv2real.htm#4C0
| Note |
|
#6. cnvMF51Acpy all <-- convert copybooks thru all steps
============= - cpy0 --> cpy1 --> cpy2 --> cpys
#7. cnvMF51A all <-- convert COBOL programs thru all steps
============ - cbl0 --> cbl1 --> cbl2 --> cbls --> cblx
#8. jcl2ksh51A all <-- convert ALL JCL thru all steps
============== proc0 --> procs, parm0 --> parms
jcl0 --> jcl1 --> jcl2 --> jcl3
#8a. jcldata51A <-- create ctl/datactl53 control file
========== - to convert JCL to scripts, jcl2 --> jcl3
| Note |
|
#9. cp jcl3/* jcls <-- copy converted JCL/scripts to jcls (in $PATH)
============== - only for our demos, not for your conversions
(will copy 1 at a time as you test/debug)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
/home/userxx/testlibs :---ctl/ : :---add/ : 245 dummy_readme :---ctl/ : 4612 cnvcob5.tbl <-- COBOL conversion search/replace table : 2238 cobdirectives <-- Micro Focus COBOL compile Directives : 0 cobfil55bI.dat <-- COBOL datafile info (initially empty) : 3072 cobfil55bI.idx : 0 datacat52I.dat <-- LISTCAT datafile info (initially empty) : 3072 datacat52I.idx : 917 datamisc51 <-- you can edit with misisng record sizes : 1334 dataxls51.psv <-- file of record sizes from Excel spread sheet : 2586 extfh.cfg <-- Micro Focus COBOL Extended File Handler config :20055 jclunixop51 <-- JCL converter control-file : 662 utilities
/home/userxx/testlibs
:---ctl/
: :---add/
: 245 dummy_readme
: 1141 datamisc51 <-- add info missing on 1st convert
: 1598 dataxls51 <-- add info from Excel spreadsheet
: 2496 datactl53a <-- copied/renamed from ../datactl53I.dat
:---ctl/
: 4612 cnvcob5.tbl
: 2238 cobdirectives
: 1292 cobfil55a <-- COBOL datafile info stored by cnvMF51A
: 1280 cobfil55aI.dat - to Identify ORG Line Seqntl (print) files
: 3072 cobfil55aI.idx
: 1292 cobfil55b <-- COBOL datafile info stored by cnvMF51A
: 1280 cobfil55bI.dat - to supply record-sizes to JCL converter
: 3072 cobfil55bI.idx
: 0 dataadd51 <-- datafile info from all ctlfiles in add/...
: 0 dataadd52I.dat - info added to datactl53I (initially empty)
: 3072 dataadd52I.idx
: 0 datacat52 <-- datafile info from LISTCAT cat2/...
: 0 datacat52I.dat - info added to datactl53I (initially empty)
: 3072 datacat52I.idx
: 1846 datactl53 <-- datafile info for JCL converter (sequential)
: 2496 datactl53I.dat - Indexed file version used by JCL converter
: 3072 datactl53I.idx
: 2074 datajcl51 <-- datafile info extracted from all jcl2/...
: 2074 datajcl52 - 1st input to create datactl53 for JCL converter
: 2586 extfh.cfg
: 207 gdgctl51 <-- initial GDG control file created by jcl2ksh51A
- copy to $RUNDATA/ctl, modify no of generations
& load Indexed file for JCL/script executions
:20055 jclunixop51
: 662 utilities
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
You must perform COBOL conversion prior to JCL conversion, because it stores control file ctl/cobfil55b with record size info for the JCL converter. The COBOL conversion was performed by a do everything script 'cnvMF51A' which was 1st documented on page '3E1' & most recently on page '3B1'.
cnvMF51Acpy all <-- convert COBOL copybooks thru all steps ============= - cpy0 --> cpy1 --> cpy2 --> cpys
cnvMF51A all <-- convert COBOL programs thru all steps ============ - cbl0 --> cbl1 --> cbl2 --> cbls --> cblx
'cnvMF51A' - includes following 6 steps:
- see listing at JCLcnv2real.htm#4B2
Here is asample of cobfile55b, created from 4 of the COBOL programs used for the test/demo conversions in this documentation. You will later see some of these record-sizes used to update ctl/datactl53I.
*cobfil55 COBOL File Info Dir=cbl1 2015/03/08_17:48:23 *progname.cbl DDname OAL open recsz copybook.cpy FDname Writes Advances * 1 2 3 4 5 6 7 *123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890 *========================================================================================== car200.cbl saledtl SS I 64 saledtl.cpy saledtl car200.cbl custmas IR I 256 custmas.cpy custmas car200.cbl salelst L L O 120 sdline.cpy salelst 2 2 cgl100.cbl acctmas SS I 128 acctmas cgl100.cbl actlist L L O 120 actlist 1 1 cgl200.cbl glmsold RS I 128 glmsold cgl200.cbl glmsnew RS O 128 glmsnew 1 * cgl200.cbl gltrans RS I 80 gltrans car100.cbl custmas SS I 256 custmas.cpy custmas car100.cbl nalist S L O 120 nalist 2 2
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#3. jcl2ksh51A all <-- convert ALL JCL components thru all steps
============== - see listing at JCLcnv2real.htm#4A1
- includes folowing functions:
#3a. cleanup proc0 --> procs, parm0 --> parms, jcl0 --> jcl1
#3b. proc expand jcl1 --> jcl2
#3c. jcldata51A <-- create ctl/datactl53 control file
========== combining info from JCL, COBOL, LISTCAT, etc
#3d. jclxx51 jcl2 jcl3 <-- convert JCL to Korn shell scripts
================= jcl2 --> jcl3
The do everything script 'jcl2ksh51A' calls script 'jcldata51A' to create control file ctl/datactl53I.dat/.idx for use by 'jclxx51' (JCL to script).
'jcldata51A' - see listing at JCLcnv2real.htm#4C1
- includes following 6 steps:
AR.CUSTOMER.MASTER rca=_____ rcs=_____ typ=RSF key=___,___ job=jar100 prg=CAR100 AR.CUSTOMER.MASTER.INDEXED rca=_____ rcs=_____ typ=RSF key=___,___ job=jar200 prg=CAR200 AR.CUSTOMER.NAMEADRS.LIST100 rca=00133 rcs=00133 typ=RSF key=___,___ job=jar100 prg=CAR100 AR.SALES.ITEMS rca=_____ rcs=_____ typ=RSF key=___,___ job=jar200 prg=SORT AR.SALES.LIST rca=00133 rcs=00133 typ=RSF key=___,___ job=jar200 prg=CAR200 GL.ACCOUNT.ACNTLIST(+1) rca=00133 rcs=00133 typ=RSF key=___,___ job=jgl100 prg=CGL100 GL.ACCOUNT.MASTER(+1) rca=00128 rcs=00128 typ=RSF key=___,___ job=jgl320 prg=IDCAMS GL.ACCOUNT.MASTER(0) rca=_____ rcs=_____ typ=RSF key=___,___ job=jgl200 prg=CGL200 GL.ACCOUNT.TRAN1 rca=_____ rcs=_____ typ=RSF key=___,___ job=jgl200 prg=SORT GL.ACCOUNT.TRANS(+1) rca=00080 rcs=00080 typ=RSF key=___,___ job=jgl230 prg=SORT
ar.customer.master rca=_____ rcs=_____ typ=RSF key=___,___ job=jar100 prg=CAR100 ar.customer.master.indexed rca=_____ rcs=_____ typ=RSF key=___,___ job=jar200 prg=CAR200 ar.customer.nameadrs.list100 rca=00133 rcs=00133 typ=RSF key=___,___ job=jar100 prg=CAR100 ar.sales.items rca=_____ rcs=_____ typ=RSF key=___,___ job=jar200 prg=SORT ar.sales.list rca=00133 rcs=00133 typ=RSF key=___,___ job=jar200 prg=CAR200 gl.account.acntlist_ rca=00133 rcs=00133 typ=RSF key=___,___ job=jgl100 prg=CGL100 gl.account.master_ rca=00128 rcs=00128 typ=RSF key=___,___ job=jgl320 prg=IDCAMS gl.account.master_ rca=_____ rcs=_____ typ=RSF key=___,___ job=jgl200 prg=CGL200 gl.account.tran1 rca=_____ rcs=_____ typ=RSF key=___,___ job=jgl200 prg=SORT gl.account.trans_ rca=00080 rcs=00080 typ=RSF key=___,___ job=jgl230 prg=SORT
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The JCL conversion was 1st documented on page '3F1' & recently on page '3B1'.
#3. jcl2ksh51A all <-- convert ALL JCL thru multiple steps
============== proc0 --> procs, parm0 --> parms, jcl0 --> jcl1 --> jcl2
| Note |
|
#3d. jclxx51 jcl2 jcl3 <-- convert all jcl2/* to Korn shell scripts in jcl3/...
================= - no need to jclxx51 if you just ran jcl2ksh51a
- might run jclxx51 to reconvert only without repeating
cleanup procs,parms,jcls,& proc expansion
'jcl2ksh51A' calls 'jclxx51' which in turn calls the actual JCL converter C program 'jcluni51' as follows:
jclunix51 jcl2/xxx.jcl jcl3/xxx.ksh ctl/datactl53I.dat/.idx ctl/cobfil55b.dat/.idx ===================================================================================
jclunix51 reads both ctl/datactl53I & cobfil55bI to get record-sizes & file-types (if mising in JCL SORTs,IDCAMS,IEBGENERs,etc). It also updates ctl/datactl53I if it finds info in cobfil55bI that was not in datactl53I.
Here is the BEFORE & AFTER listing of ctl/datactl53I showing record-size info transferred from cobfil55bI (listed on previous page) & used to update ctl/datactl53I.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here are listings of the 2 versions of the JCL converter control file (before & after running the 1st JCL conversion).
| ctl/datactl53 |
|
ar.customer.master rca=_____ rcs=_____ gdg= typ=RSF key=___,___ job=jar100 prg=CAR100 ar.customer.master.indexed rca=_____ rcs=_____ gdg= typ=RSF key=___,___ job=jar200 prg=CAR200 ar.customer.nameadrs.list100 rca=00133 rcs=00133 gdg= typ=RSF key=___,___ job=jar100 prg=CAR100 ar.sales.items rca=_____ rcs=_____ gdg= typ=RSF key=___,___ job=jar200 prg=SORT ar.sales.list rca=00133 rcs=00133 gdg= typ=RSF key=___,___ job=jar200 prg=CAR200 gl.account.acntlist_ rca=00133 rcs=00133 gdg= typ=RSF key=___,___ job=jgl100 prg=CGL100 gl.account.master_ rca=00128 rcs=00128 gdg= typ=RSF key=___,___ job=jgl320 prg=CGL200 gl.account.tran1 rca=_____ rcs=_____ gdg= typ=RSF key=___,___ job=jgl200 prg=SORT gl.account.trans_ rca=00080 rcs=00080 gdg= typ=RSF key=___,___ job=jgl230 prg=SORT
| ctl/datactl53I.dat |
|
ar.customer.master rca=00256 rcs=00256 gdg= typ=RSF key=___,___ job=jar100 prg=CAR100 ar.customer.master.indexed rca=00256 rcs=00256 gdg= typ=RSF key=___,___ job=jar200 prg=CAR200 ar.customer.nameadrs.list100 rca=00120 rcs=00120 gdg= typ=LSTt key=___,___ job=jar100 prg=CAR100 ar.sales.items rca=00064 rcs=00064 gdg= typ=RSF key=___,___ job=jar200 prg=SORT ar.sales.list rca=00120 rcs=00120 gdg= typ=LSTt key=___,___ job=jar200 prg=CAR200 gl.account.acntlist_ rca=00120 rcs=00120 gdg= typ=LSTt key=___,___ job=jgl100 prg=CGL100 gl.account.master_ rca=00128 rcs=00128 gdg= typ=RSF key=___,___ job=jgl320 prg=CGL200 gl.account.tran1 rca=00080 rcs=00080 gdg= typ=RSF key=___,___ job=jgl200 prg=SORT gl.account.trans_ rca=00080 rcs=00080 gdg= typ=RSF key=___,___ job=jgl230 prg=SORT
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
If available, the mainframe LISTCAT reports can be transferred to unix & the data file info (record-sizes, file-types, Indexed keys) can be extracted & merged into the JCL converter control file to greatly improve JCL conversions.
#1. login yourself --> /home/userxx
#2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3. mkdir cat0 cat1 cat2 <-- make subdirs for LISTCAT files processing
#4. cd cat0
#5. FTP LISTCAT reports from mainframe to unix
- may be multiple files, names do not matter, but must be valid for unix
#6. cdl <-- change back to working directory (above cat0,cbl0,jcl0,etc)
#7. <-- change names as necessary for unix #7a. renameL cat0 - rename to Lower case for unix #7b. renameB2_ cat0 - change any embedded blanks to underscores #7c. renameD2_ cat0 - change any '$' dollar signs to '_' underscores #7d. rename-QQ cat0 - remove any 'Quotes' #7e. rename-PP cat0 - remove any (Parenthesis)
#8. catdata50 all <-- extract datafile info from LISTCAT reports
=============
cat0 -------> cat1 ---------> cat2 --------> ctl/datacat51 --> ctl/datacat52I
catdata51 catdata52 catcat51 uvcp
| Note |
|
We had a problem at 1 site - the record sizes were misisng from the LISTCAT report for NON-VSAM files - but the mainframe person was able to create an additional report with LRECLs which we converted to the keyword format & combined with the other files - see details on page '3D1'.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
# listcat001 - test file for Vancouver Utilities JCL & DATA conversion # - see www.uvsoftware.ca/datacnv1.htm#Part_6 # or www.uvsoftware.ca/jclcnv2real.htm#3C2 # 'CLUSTER's for a few files extracted from mainframe LISTCAT report # - omitting many items not relevant to desired information # - selecting: KEYLEN, RKP, AVGLRECL, MAXLRECL, RECTOTAL, LIMIT (for GDGs)
0CLUSTER ------- AR.CUSTOMER.MASTER
KEYLEN-----------------0 AVGLRECL-----------256 BUFSPACE---------14366
RKP--------------------0 MAXLRECL-----------256 EXCPEXIT--------(NULL)
REC-TOTAL-------------32 SPLITS-CI------------0 EXCPS--------------495
0CLUSTER ------- AR.CUSTOMER.MASTER.INDEXED
ASSOCIATIONS
DATA-----AR.CUSTOMER.MASTER.INDEXED.DATA
INDEX----AR.CUSTOMER.MASTER.INDEXED.INDEX
KEYLEN-----------------6 AVGLRECL-----------256 BUFSPACE---------14366
RKP--------------------0 MAXLRECL-----------256 EXCPEXIT--------(NULL)
REC-TOTAL-------------32 SPLITS-CI------------0 EXCPS--------------495
0CLUSTER ------- AR.SALES.ITEMS
KEYLEN-------0 AVGLRECL--------64 BUFSPACE------4096 CISIZE--------2048
RKP----------0 MAXLRECL--------64 EXCPEXIT------(NULL) CI/CA----------270
REC-TOTAL---20 SPLITS-CI------0 EXCPS----------8
0GDG BASE ------ GL.ACCOUNT.MASTER
KEYLEN-------0 AVGLRECL-------128 BUFSPACE------4096 CISIZE--------2048
RKP----------0 MAXLRECL-------128 EXCPEXIT------(NULL) CI/CA----------270
LIMIT------------------5 SCRATCH NOEMPTY
ASSOCIATIONS
NONVSAM--GL.ACCOUNT.MASTER.G0005V00
NONVSAM--GL.ACCOUNT.MASTER.G0006V00
NONVSAM--GL.ACCOUNT.MASTER.G0007V00
0CLUSTER ------- GL.ACCOUNT.TRANS
KEYLEN-------0 AVGLRECL--------80 BUFSPACE------4096 CISIZE--------2048
RKP----------0 MAXLRECL--------80 EXCPEXIT------(NULL) CI/CA----------270
LIMIT------------------6 SCRATCH NOEMPTY
ASSOCIATIONS
NONVSAM--GL.ACCOUNT.TRANS.G0001V00
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
AR.CUSTOMER.MASTER rca=00256 rcs=00256 typ=RSF cnt=00000032 AR.CUSTOMER.MASTER.INDEXED rca=00256 rcs=00256 typ=IDXf1 cnt=00000032 key=000,006 AR.SALES.ITEMS rca=00064 rcs=00064 typ=RSF cnt=00000020 GL.ACCOUNT.MASTER_ rca=00000 rcs=00000 typ=RSF gdg=003 GL.ACCOUNT.TRANS rca=00080 rcs=00080 typ=RSF
ar.customer.master rca=00256 rcs=00256 typ=RSF cnt=00000032 ar.customer.master.indexed rca=00256 rcs=00256 typ=IDXf1 cnt=00000032 key=000,006 ar.sales.items rca=00064 rcs=00064 typ=RSF cnt=00000020 gl.account.master_ rca=00000 rcs=00000 typ=RSF gdg=003 gl.account.trans rca=00080 rcs=00080 typ=RSF
ar.customer.master rca=00256 rcs=00256 gdg= typ=RSF ar.customer.master.indexed rca=00256 rcs=00256 gdg= typ=IDXf1 key=000,006 ar.sales.items rca=00064 rcs=00064 gdg= typ=RSF gl.account.master_ rca=_____ rcs=_____ gdg=003 typ=RSF gl.account.trans rca=00080 rcs=00080 gdg= typ=RSF
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We had a problem at 1 site - the record sizes were misisng from the LISTCAT report for NON-VSAM files - but the mainframe person was able to create an additional report which showed LRECL's as follows:
DATA SET NAME VOLUME ORG RECFM LRECL BLKSZ
PGAPO.ABE.EKSTADA.VLK AGS102 PO FB 80 6160
PGAPO.DEHUF.ABE.ABEBUCH.ONLINE AGS039+ PS FB 350 5950
PGAPO.DEHJW.GAPOMYEB.MGYART AGS073+ PS FB 4000 8000
PGAPO.EDSKW.GAPO41F2.KAGS416.LISTE2.RE AGS043+ PS FBA 133 6118
PGAPO.EDSKW.GAPO41F2.KAGS420.AUSGABE.RE AGS102+ PS FB 14150 14150
PGAPO.DEHBU.GAPOAYK2.FTP.KRCOPY.G0398V00 AGS065+ PS FB 80 1600
I then wrote a uvcopy job to convert the LRECL listing into the keyword format used by the Vancouver Utilities JCL & DATA conversion jobs.
uvcopy lrecl0cat2,fili1=ctl/lrecl0,filo1=cat2/lrecl2 ===================================================== - convert mainframe LRECL listing to VU keyword format
pgapo.abe.ekstada.vlk rca=00080 rcs=00080 typ=RSF pgapo.dehuf.abe.abebuch.online rca=00350 rcs=00350 typ=RSF pgapo.dehjw.gapomyeb.mgyart rca=04000 rcs=04000 typ=RSF pgapo.dehpd.gapoaff5.kags416.liste2 rca=00133 rcs=00133 typ=RSF pgapo.edskw.gapoayf2.kags420.ausgabe.re rca=14150 rcs=14150 typ=RSF pgapo.dehbu.gapoayk2.ftp.krcopy_ rca=00080 rcs=00080 typ=RSF gdg=___
Note - GDG files recognized by last node '.G####V00'
- replaced by trailing '_' ID for UV GDG file & gdg=___ appended
This job is part of script 'catdata50' which calls following jobs: catdata51, catdata52, lrecl0cat2 (this job), cp, catcat51,& uvcp
catdata51 - convert mainframe LISTCAT report to keyword format
catdata52 - translate filenames to lower case & convert GDG files ot VU format
*lrecl0cat2 - converts LRECL mainframe report to keyword format cat2/lrecl2
cp - copies cat2/lrecl2 to ctl/add for JCL converter control files
catcat51 - combines all files in cat2/... into ctl/datacat51
uvcp - loads ctl/datacat52I, to assist JCL conversion
- ctl/datacat52I also input to DATA conversion job 'ctlcat12'
JCL converter script jcl2ksh51A calls script jcldata51, which calls uvcopy job ctldata53, which loads ctl/datactl53I (for the JCL converter) from various inputs including ctl/datacat52I
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
If after 1st 2 conversions, you notice missing record-sizes or incorrect file-types for some files, you can add your own-coded control file in the ctl/add/ subdir, knowing that the keyword info from all files in ctl/add/* are merged into the JCL converter control file on the next conversion. Here is the sample/demo supplied in $UV/ctl/datamisc51.
# datamisc51 - miscellaneous data file info for JCL conversion
# - demo file for www.uvsoftware.ca/jclcnv2real.doc#Part_2
#
# Must be stored in $RUNLIBS/ctl/add/...
# But could have any name desired BECAUSE:
# - uvcopy catcat51 (part of jcldata51A script, called by jcl2ksh51A)
# - will combine all files found in ctl/add/... into ctl/dataadd51
# which is loaded into Indexed file ctl/dataadd51I.dat/.idx
# which is looked up by uvcopy ctldata53 to merge any keyword data found
# into ctl/datactl53, loaded into ctl/datactl53I.dat/.idx for JCL converter
# These comment lines ('#' in column 1) will be dropped by catcat51
# Here are a few lines keyword info for the demo files
#
ar.sales.items rca=00064 rcs=00064 typ=RST
gl.account.acntlist_ rca=00132 rcs=00132 typ=LSTt2 gdg=003
gl.account.trans_ rca=00080 rcs=00080 typ=RST gdg=007
For our JCL conversion #2, we will copy the demo file into ctl/add/... where it will be combined with other control files on the next JCL conversion.
#1. login yourself --> /home/userxx
#2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3. cp ctl/datamisc51 ctl/add/
==========================
- copy demo manually prepared control file into ctl/add/...
to make it effective on next JCL conversion
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Some users might prefer to create am Excel spreadsheet to code missing record sizes, file types, Indexed keys, gdgs, etc. OR they may already have Excel spreadsheets that were used for the mainframe.
Here is a sample Excel spreadsheet supplied in $UV/ctl/dataxls51.psv. This is the pipe|delimited version after being exported from the Excel spreadsheet.
DSN|Volume|Organization|Reclen|Block|RecFormat|Allocated|USED|Extents|created|Referenced ar.customer.master|VOL001|PS|256|25600|FB|800|50|1|9/30/2014|3/14/2015| ar.sales.items|VOL002|PS|64|6400|FB|750|25|1|8/25/2014|2/17/2015| ar.customer.nameadrs.list100|VOL003|PS|132|1320|FB|750|25|1|8/25/2014|2/17/2015|
For this spreadsheet, the only item applicable to unix was the record-size. We made a uvcopy job 'xlsdata51' to extract the data into our required format For example:
ar.customer.master rca=00256 rcs=00256 data=ps__fb_ ar.sales.items rca=00064 rcs=00064 data=ps__fb_ ar.customer.nameadrs.list100 rca=00132 rcs=00132 data=ps__fb_
The Output file will be in $RUNLIBS/ctl/add/ where it will be combined with any other control files on the next JCL conversion
See the uvcopy job 'xlsdata51' listed at JCLcnv2real.htm#4C7 You might modify it depending on the format of your Excel spreadsheet. You can run as follows:
#1. login yourself --> /home/userxx
#2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3. uvcopy xlsdata51,fili1=ctl/dataxls51.psv,filo1=ctl/add/xlsdata51
================================================================
#3a. uvcopy xlsdata51 <-- same as above, files default as shown
================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. login yourself --> /home/userxx
#2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3a. mv jcl3 jcl3.old <-- save prior conversions by changing dirname
================ - so we can use alldiff2 below to see differences
#3b. mkdir jcl3 <-- make new subdir for re-conversions
==========
#4. jcl2ksh51A all <-- convert ALL JCL thru all steps
============== proc0 --> procs, parm0 --> parms
jcl0 --> jcl1 --> jcl2 --> jcl3
I highly recommend this procedure to see the differences that re-converting makes - to ensure you changed what you intended & did not cause any unintended changes.
#5a. alldiff2 jcl3.old jcl3 <-- create diff report in tmp/jcl3.dif
======================
#5b. vi tmp/jcl3.dif <-- investigate diff report
===============
--> :g/ver:/d <-- drop diffs caused by JCL converter version diff
--> :g/logmsg/d <-- drop diffs caused by logmsg lines
| Note |
|
Dropping the version, logmsg, & other non-relevant diffs lets you concentrate on the important differences, but can cause many empty difference files (where only diff was the version). I suggest you use the 'uvcopy diffdrop2' job to eliminate these lines:
#6. uvcopy diffdrop2,fili1=tmp/jcl3.dif,filo1=tmp/jcl3.dif2
=======================================================
#6a. uvcopy diffdrop2 <-- same as above (files default as shown)
================
#7. vi tmp/jcl3.dif2 <-- inestigate shorter report
================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The previous page gave the instructions to run script 'alldiff2' (after reconverting the JCL) to create a differences report to prove that the reconversions changed what was intended with no unintended consequences. Here we will illustrate 3 versions of the alldiff2 report:
To save space, we will show results for only 4 of the 9 demo JCL/scripts.
39c39 < exit 0 #ver:20150313 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 --- > exit 0 #ver:20150314 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 diff file# 2 - jcl3.old/... vs jcl3/jar100.ksh
31c31 < logmsg2 "Executing--> uvsort \"fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTO..." --- > logmsg2 "Executing--> uvsort \"fili1=$SORTIN,typ=RST,rcs=64,filo1=$SORTO..." 33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ --- > uvsort "fili1=$SORTIN,typ=RST,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ 70c70 < exit 0 #ver:20150313 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 --- > exit 0 #ver:20150314 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 diff file# 3 - jcl3.old/... vs jcl3/jar200.ksh
47c47 < exit 0 #ver:20150313 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 --- > exit 0 #ver:20150314 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 diff file# 4 - jcl3.old/... vs jcl3/jgl100.ksh
38c38 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RST,rcs=80\ 76c76 < exit 0 #ver:20150313 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 --- > exit 0 #ver:20150314 a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1 diff file# 5 - jcl3.old/... vs jcl3/jgl200.ksh 9 different of 9 files compared jcl3.old to jcl3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#5b. vi tmp/jcl3.dif <-- investigate diff report
===============
--> :g/ver:/d <-- drop diffs caused by JCL converter version diff
--> :g/logmsg/d <-- drop diffs caused by logmsg lines
39c39 --- diff file# 2 - jcl3.old/... vs jcl3/jar100.ksh
31c31 --- 33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ --- > uvsort "fili1=$SORTIN,typ=RST,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ 70c70 --- diff file# 3 - jcl3.old/... vs jcl3/jar200.ksh
47c47 --- diff file# 4 - jcl3.old/... vs jcl3/jgl100.ksh
38c38 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RST,rcs=80\ 76c76 --- diff file# 5 - jcl3.old/... vs jcl3/jgl200.ksh 9 different of 9 files compared jcl3.old to jcl3
33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ --- > uvsort "fili1=$SORTIN,typ=RST,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ diff file# 3 - jcl3.old/... vs jcl3/jar200.ksh
38c38 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RST,rcs=80\ diff file# 5 - jcl3.old/... vs jcl3/jgl200.ksh
33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RST,rcs=80\ diff file# 6 - jcl3.old/... vs jcl3/jgl230.ksh 9 different of 9 files compared jcl3.old to jcl3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Page '3F1' re-converted using 'jcl2ksh51A' (do everything script).
#4. jcl2ksh51A all <-- convert ALL JCL thru all steps
============== proc0 --> procs, parm0 --> parms
jcl0 --> jcl1 --> jcl2 --> jcl3
But jcl2ksh51A repeats cleanups & proc expands that are unnecessary if we are only changing control files. If you have hundreds of JCLs, PROCs,& PARMs, you can save time by replacing #4 above as shown below with steps #4a, #4b,& #4c.
#1. login yourself --> /home/userxx #2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#4a. jcldata51A <-- re-create ctl/datactl53 control file
========== including LISTCAT info
#4b. jclxx51 jcl2 jcl3 <-- re-convert JCL to Korn shell scripts
=================
#4c. cp ctl/datactl53I.dat ctl/add/datactl53a
========================================
- save updated JCL converter control file for future re-conversions
'jcl2ksh51A' (reconvert everything script) includes #4c. Since we are replacing jcl2ksh51A with jcldata51A & jclxx51, we also do this manually.
We do this because the converter updates the Indexed file datactl53I.dat/.idx with any datafile info supplied by other sources such as COBOL, LISTCAT, dataxls (spreadsheets), datamisc (misc edited info). We copy the file into ctl/add/... because all files there are combined for following re-conversions.
The jcl2ksh51A script renames as shown above on #4c, but there might be some advantage in renaming it as follows (using suffix 'b' vs 'a').
#4c2. cp ctl/datactl53I.dat ctl/add/datactl53b
=======================================*
This allows you to make manual updates to ctl/add/datactl53b & not lose them on reconversion - because ctl/add/datactl53a is overwritten by reconversion, but any added info in ctl/add/datactl53a will be merged into new control files & especially 'dataclt53I.dat/.idx' used by the JCL converter.
#4d. vi ctl/add/datactl53b <-- manual update
=====================
- add record-sizes, change file-types, etc as desired
#5. cp jcl3/* jcls <-- copy converted JCL/scripts to jcls (in $PATH)
============== - only for our demos, not for your conversions
(will copy 1 at a time as you test/debug)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
ar.customer.master rca=00256 rcs=00256 gdg=___ typ=RSF key=___,___ job=jar100 prg=CAR100 ar.customer.master.indexed rca=00256 rcs=00256 gdg=___ typ=RSF key=___,___ job=jar200 prg=CAR200 ar.customer.nameadrs.list100 rca=00120 rcs=00120 gdg=___ typ=LSTt key=___,___ job=jar100 prg=CAR100 ar.sales.items rca=00064 rcs=00064 gdg=___ typ=RSF key=___,___ job=jar200 prg=SORT ar.sales.list rca=00120 rcs=00120 gdg=___ typ=LSTt key=___,___ job=jar200 prg=CAR200 gl.account.acntlist_ rca=00120 rcs=00120 gdg=___ typ=LSTt key=___,___ job=jgl100 prg=CGL100 gl.account.master_ rca=00128 rcs=00128 gdg=___ typ=RSF key=___,___ job=jgl320 prg=CGL200 gl.account.tran1 rca=00080 rcs=00080 gdg=___ typ=RSF key=___,___ job=jgl200 prg=SORT gl.account.trans_ rca=00080 rcs=00080 gdg=___ typ=RSF key=___,___ job=jgl230 prg=SORT
ar.customer.master rca=00256 rcs=00256 gdg=___ typ=RSF key=___,___ job=jar100 prg=CAR100 ar.customer.master.indexed rca=00256 rcs=00256 gdg=___ typ=IDXf1 key=000,006 job=jar200 prg=CAR200 ar.customer.nameadrs.list100 rca=00120 rcs=00120 gdg=___ typ=LSTt key=___,___ job=jar100 prg=CAR100 ar.sales.items rca=00064 rcs=00064 gdg=___ typ=RST key=___,___ job=jar200 prg=SORT ar.sales.list rca=00120 rcs=00120 gdg=___ typ=LSTt key=___,___ job=jar200 prg=CAR200 gl.account.acntlist_ rca=00120 rcs=00120 gdg=003 typ=LSTt key=___,___ job=jgl100 prg=CGL100 gl.account.master_ rca=00128 rcs=00128 gdg=005 typ=RSF key=___,___ job=jgl320 prg=CGL200 gl.account.tran1 rca=00080 rcs=00080 gdg=___ typ=RSF key=___,___ job=jgl200 prg=SORT gl.account.trans_ rca=00080 rcs=00080 gdg=007 typ=RST key=___,___ job=jgl230 prg=SORT
1a. ar.customer.master.indexed now coded corectly as typ=IDXf1 & key=000,006
- picked up from the LISTCAT reports (used on 2nd convert but not 1st.
1b. 3 of the GDG files now have the number of generations coded - picked up from the LISTCAT reports.
1c. ar.sales.items & gl.account.trans changed 'typ=RSF' to 'typ=RST'.
- picked up from ctl/datamisc51 (used on 2nd convert but not 1st).
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We will illustrate changing option 'f0/f1' in ctl/jclunixop51 to change default file typ=RSF to typ=LST for sort & copy utiltiies. We can also change individual file types & record-sizes in ctl/datactl51 & reloading Indexed file ctl/datactl51I & reconverting.
The JCL converter control file is listed at JCLcnv2real.htm#4C1 Here is the options (line 8) before & after changing option 'f0' to 'f1'
jclunixop51:a2b0c0d1e2f0g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1
==========**==========================================
jclunixop51:a2b0c0d1e2f1g1h1i15j0k15l1m4n3o8p0q0r0s2t1u1v0w0x0y1z1
Here are descriptions for option 'f' from lines 47-49 of ctl/jclunixop51
# f0 - default file typ=RSF (allows packed/binary) # f1 - default uvsort/uvcp file typ=LST # f2 - default uvsort/uvcp file typ=LSTt
You could change to 'f1' if you have very few files with packed/binary. Conversions of SORT, IEBGENER,etc to uvsort & uvcp would then code file types as 'typ=LST' vs 'typ=RSF'. The advantage is that you could then use unix utilities that equire LineFeeds (vi,more,cat,lp) on your datafiles. You could manually change some uvsort/uvcp 'typ=LST' back to 'typ=RSF' for a few files with packed/binary.
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#3. vi $RUNLIBS/ctl/jclunixop51 <-- change option 'f0' to 'f1'
=========================== (as shown above)
#4a. mv jcl3 jcl3.old <-- save existing conversions old/new diff check
================
#4b. mkdir jcl3 <-- make new subdir to reveive re-conversions
==========
#5. jclxx51 jcl2 jcl3 <-- reconvert all JCL jcl2/* --> jcl3/...
================= (after option changes)
#6. alldiff2 jcl3.old jcl3 <-- run 'diff' on all files in old/new subdirs
====================== - writes report to tmp/jcl3.dif
#6a. vi tmp/jcl3.dif <-- check differences as expected
=============== & no unintended changes
- see sample report on next page below
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ --- > uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=LST,rcs=64\ diff file# 3 - jcl3.old/... vs jcl3/jar200.ksh
30c30 < uvcp "fili1=$SYSUT1,typ=RSF,rcs=128,filo1=$SYSUT2,typ=RSF,rcs=128" --- > uvcp "fili1=$SYSUT1,typ=RSF,rcs=128,filo1=$SYSUT2,typ=LST,rcs=128" 59c59 < uvcp "fili1=$TEMPGLM,typ=RSF,rcs=80,filo1=$ACNTMAS,typ=RSF,rcs=128"; --- > uvcp "fili1=$TEMPGLM,typ=LST,rcs=80,filo1=$ACNTMAS,typ=RSF,rcs=128"; diff file# 7 - jcl3.old/... vs jcl3/jgl320.ksh
#7. cp jcl3/* jcls <-- copy re-converted jcl/scripts for executions
============== - 'jcls' is in the PATH defined in our profile
| Note |
|
#7a. cp jcl3/jobxx.ksh jcls
======================
- copy 1 at a time when you are ready to debug each JCL/script
#8. joblog jar200.ksh <-- might rerun jar200 to check changes
================= in file type output by uvsort typ=LST
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
These 3 pages 3G1-3G3 are older & perhaps redundant to page '3D1' coding ctl/add/datamisc51 with changes to file info for selected files & to page '3E4' which 1st suggested copying ctl/datactl53I.dat to ctl/add/datactl53b and making changes there for future conversions.
The JCL converter script jcl2ksh51A extracts datafile info (types,rec-sizes,etc) from all JCLs & creates an Indexed control file ctl/datactl53I that is used by the JCL converter jclunix51.c to supply datafile info to JCL jobs that may not define datafile types & record-sizes required (by uvsort,uvcp,etc). Note that the mainframe could get file info from the VTOC & omit from JCL, but file info must be defined on uvsort & uvcp steps.
ar.customer.master cpy= rcs=_____ gdg= typ=RSF data= job=jar100 prg=CAR100 ar.customer.master.indexed cpy= rcs=_____ gdg= typ=RSF data= job=jar200 prg=CAR200 ar.customer.nameadrs.list100 cpy= rcs=00133 gdg= typ=RSF data= job=jar100 prg=CAR100 ar.parmslib@parm200s cpy= rcs=_____ gdg= typ=RSF data= job=jar200 prg=SORT ar.sales.items cpy= rcs=00064 gdg= typ=RSF data= job=jar200 prg=SORT ar.sales.list cpy= rcs=00133 gdg= typ=RSF data= job=jar200 prg=CAR200 gl.account.acntlist_ cpy= rcs=00133 gdg= typ=RSF data= job=jgl100 prg=CGL100 gl.account.master_ cpy= rcs=00128 gdg= typ=RSF data= job=jgl320 prg=CGL200 gl.account.tran1 cpy= rcs=00080 gdg= typ=RSF data= job=jgl200 prg=SORT gl.account.trans_ cpy= rcs=00080 gdg= typ=RSF data= job=jgl230 prg=SORT gl.parms@pgl200s1 cpy= rcs=_____ gdg= typ=RSF data= job=jgl200 prg=SORT
The JCL conversion script jcl2ksh51A saves a text version of ctl/datactl53I in ctl/add/datactl53a. You could copy this to ctl/add/datactl53b & update as desired. The next run of jcl2ksh51A will include any new info supplied in the recreated ctl/datactl53I.
ar.customer.master cpy= rcs=00256 gdg= typ=RSF data=p job=jar100 prg=CAR100 ar.customer.master.indexed cpy= rcs=00256 gdg= typ=RSF data=p job=jar200 prg=CAR200 ar.customer.nameadrs.list100 cpy= rcs=00133 gdg= typ=LSTt data= job=jar100 prg=CAR100 ar.parmslib@parm200s cpy= rcs=_____ gdg= typ=RSF data= job=jar200 prg=SORT ar.sales.items cpy= rcs=00064 gdg= typ=LST data= job=jar200 prg=SORT ar.sales.list cpy= rcs=00133 gdg= typ=LSTt data= job=jar200 prg=CAR200 gl.account.acntlist_ cpy= rcs=00133 gdg=05 typ=LSTt data= job=jgl100 prg=CGL100 gl.account.master_ cpy= rcs=00128 gdg=05 typ=RSF data=p job=jgl320 prg=CGL200 gl.account.tran1 cpy= rcs=00080 gdg= typ=LST data= job=jgl200 prg=SORT gl.account.trans_ cpy= rcs=00080 gdg=05 typ=LST data= job=jgl230 prg=SORT gl.parms@pgl200s1 cpy= rcs=_____ gdg= typ=RSF data= job=jgl200 prg=SORT
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
jcl2ksh51A all <-- script to convert ALL JCL thru all steps ============== jcl0 --> jcl1 --> jcl2 --> jcl3
This assumes that we have already converted all JCL as shown above (see instructions & console log beginning on page '3F1'). After some initial testing, we may wish to modify the data control files & re-convert.
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#3. cp ctl/add/datactl53a ctl/add/datactl53b
========================================
- copy data ccontrol file saved by 1st conversion
for modify & inclusion in subsequent conversions
#4. vi ctl/add/datactl53b
=====================
- see sample modifications on page above
#5. rm ctl/add/datactl53a
=====================
- OK to remove the '53a' file created by 1st conversion
- it will be recreated by reconversion script 'jcl2ksh51A' below
(containing the info from the '53b' file you have edited)
#6a. mv jcl3 jcl3.old <-- save existing conversions old/new diff check
================
#6b. mkdir jcl3 <-- make new subdir to reveive re-conversions
==========
#7. jcl2ksh51A all <-- re-convert all JCL thru all steps
============== jcl0 --> jcl1 --> jcl2 --> jcl3
- to include new data control file info
#8. alldiff2 jcl3.old jcl3 <-- run 'diff' on all files in old/new subdirs
====================== - writes report to tmp/jcl3.dif
#8a. vi tmp/jcl3.dif <-- check differences as expected
=============== & no unintended changes
- see sample report on next page below
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#8. alldiff2 jcl3.old jcl3 <-- run 'diff' on all files in old/new subdirs
====================== - writes report to tmp/jcl3.dif
#8a. vi tmp/jcl3.dif <-- check differences as expected
=============== & no unintended changes
- sample report below
33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ --- > uvsort "fili1=$SORTIN,typ=LST,rcs=64,filo1=$SORTOUT,typ=RSF,rcs=64\ diff file# 3 - jcl3.old/... vs jcl3/jar200.ksh
37c37 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=LST,rcs=80,filo1=$SORTOUT,typ=LST,rcs=80\ diff file# 5 - jcl3.old/... vs jcl3/jgl200.ksh
33c33 < uvsort "fili1=$SORTIN,typ=RSF,rcs=80,filo1=$SORTOUT,typ=RSF,rcs=80\ --- > uvsort "fili1=$SORTIN,typ=LST,rcs=80,filo1=$SORTOUT,typ=LST,rcs=80\ diff file# 6 - jcl3.old/... vs jcl3/jgl230.ksh
3 different of 9 files compared jcl3.old to jcl3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here is a summary of the data control files used by the JCL converter super- script 'jcl2ksh51A'. This should help you understand the process.
| jcl2ksh51A |
|
| jcldata51A |
|
3a. ctl/datacat52I - Indexed file created by script catdata50
from multiple mainframe LISTCAT reports
- empty file created by copymvsctls (in case no LISTCAT)
3b. ctl/add/... - may be multiple sequential files in ctl/add/... subdir
- will be combined by uvcopy catcat51 & loaded into
Indexed file ctl/dataadd52I
3c. ctl/dataadd52I - Indexed file (interim) created by this script
from all additional info files in ctl/add/...
3d. $UV/ctl/add/dummy_readme supplied if no additional info
- copied to ctl/add by copymvsctls (in case no other files)
3e. additional files that you may create, name as you like, for example: ctl/datacpy52 - datafilenames & copybooknames ctl/dataedt52 - make with editor to supply missing info as desired ctl/datagdg52 - make with editor to supply missing GDG gnerations
3f. ctl/add/datactl53a - input from any prior jcl2ksh51A, converts all JCL
jcl0 --> jcl1 --> jcl2 --> jcl3
- saves updates to datactl53I made by JCL converter
for reconversions by jcl2ksh51A
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
You might want to setup an alternate set of testlibs/testdata/cnvdata. You can name them whatever you want, as long as they are defined in your profile by RUNLIBS/RUNDATA/VNCDATA. You use an alternate set of libraries & data to concentrate on a small system of JCL, COBOL,& DATA files (and make it easier to test/debug). For example, here is how you might setup alternate set testlibs1/testdata1/cnvdata1 & modify profile to match:
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. mkdir testlibs1 testdata1 cnvdata1 <-- setup alternate superdirs
==================================
#3. vi .bash_profile <-- edit profile to redefine RUNLIBS/RUNDATA/CNVDATA
#3a. vi .profile (.profile for ksh shell vs bash shell)
================ - see profile listed on page '1C1'
export RUNLIBS=$HOME/testlibs <-- existing definitions
export RUNDATA=$HOME/testdata
export CNVDATA=$HOME/cnvdata
export RUNLIBS=$HOME/testlibs1 <-- update to new definitions
export RUNDATA=$HOME/testdata1
export CNVDATA=$HOME/cnvdata1
#4. Logoff & backon to activate new definitions
#5a. cdl --> /home/testlibs1 (alias cdl='cd $RUNLIBS)
=======================
#5b. mvslibsdirs <-- script sets up 30 subdirs (see page '3B2')
===========
#6a. cdd --> /home/testdata1 (alias cdd='cd $RUNDATA)
=======================
#6b. mvsdatadirs <-- script sets up 15 subdirs (see page '3B3')
===========
#7a. cdc --> /home/cnvdata1 (alias cdl='cd $CNVDATA)
=======================
#7b. cnvdatadirs <-- script sets up 25 subdirs (see page '3B3')
===========
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Above we have setup alternate set of conversion directories testlibs1/testdata1. We might now copy over a subset of JCL/COBOL & convert/test separately. We might do this for a simpler learning experience.
#8a. cp ../teslibs/jcl0/... jcl0/ <-- select desired subset of JCL
============================
#8b. cp ../teslibs/proc0/... proc0/ <-- select desired subset of procs
==============================
#8c. cp ../teslibs/parm0/... parm0/ <-- select desired subset of parms
==============================
#9a. cp ../teslibs/cbl0/... cbl0/ <-- select desired subset of COBOL
============================
#9b. cp ../teslibs/cpy0/... cpy0/ <-- select desired subset of copybooks
============================
#10a. cnvMF51Acpy all <-- convert copybooks thru all steps
============= - see prompt/reply responses at '3E2'
#10b. cnvMF51A all <-- convert COBOL programs thru all steps
============= - see prompt/reply responses at '3E3'
#11. jcl2ksh51A all <-- convert ALL JCL thru all steps
============== jcl0 --> jcl1 --> jcl2 --> jcl3
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Script 'mvsfiles51' creates the 'essential files' report for 1 job at a time (vs 'mvsfiles5A' for all jobs in jcl2/...). Run mvsfiles51 before you attempt to debug each JCL/script to determine if you have all INPUT files required for a particular job. See complete documentation on pages '1L1' - '1L8', but here is a demo for JCL jgl230.jcl (3 steps: SORT + COBOL cgl100.cbl & cgl200.cbl).
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl --> $HOME/testlibs (/home/userxx/testlibs)
#3. mkdir xmvs <-- make subdir for temp files & final report
==========
#4. mvsfiles51 jcl2/jgl230.jcl <-- script for 1 job at a time
==========================
#5. vi xmvs/mvsfiles5 <-- inspect report of all files SORTED by FILENAME
================= - more useful to determine INPUT files required
mvsfiles5 - Insert '*' Flags beside Esential Input files - 2013/12/31 08:57:22 Jobname Step Program DDname MOD Gen Rcsz pb From To * <-----DSNname------> ===============================================================================
jgl230 003 CGL100 ACTLIST MKD- +1 133 GL.ACCOUNT.ACNTLIST_
jgl230 002 CGL200 GLMSOLD O 0 0128 p 003 * GL.ACCOUNT.MASTER_ jgl230 002 CGL200 GLMSNEW NCD- +1 128 p 003 GL.ACCOUNT.MASTER_ jgl230 003 CGL100 ACCTMAS S 0 0128 p 002 GL.ACCOUNT.MASTER_
jgl230 001 SORT SORTIN O * GL.ACCOUNT.TRAN1
jgl230 001 SORT SORTOUT NCD- +1 80 002 GL.ACCOUNT.TRANS_ jgl230 002 CGL200 GLTRANS O - +1 0080 001 GL.ACCOUNT.TRANS_
# mvsfiles7 - copy essential files from $CNVDATA/d2asc/... to $RUNDATA/data/... cp $CNVDATA/d2asc/gl.account.master_* $RUNDATA/data1 cp $CNVDATA/d2asc/gl.account.tran1 $RUNDATA/data1
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Printer files are not automatically sent to a printer as they might have been on the mainframe, but you can insert print commands into the JCL/scripts. Here is an example based on 'jar100.ksh'. See complete listing on page '2A1', and here is the last 20 lines defining the files & executing COBOL program 'car100.cbl' which creates print file data1/ar.customer.nameadrs.list100.
-------- 1st 22 lines omitted, see complete listing on page '2A1' --------
exportfile CUSTMAS data1/ar.customer.master
exportfile NALIST data1/ar.customer.nameadrs.list100 #<--NOTE print file
exportfile SYSOUT $SYOT/${jobid2}_${JSTEP}_SYSOUT_$(date +%y%m%d_%H%M%S)
#3----------------------------------------------------------------------
cobrun $ANIM $RLX/car100
#4----------------------------------------------------------------------
LCC=$?; S0010C=$LCC; ((SCC|=LCC)); ((JCC|=LCC)); S0010R=1; alias goto="";
if ((S0010C != 0))
then logmsg2 "ERR: step#$JSTEP car100 abterm $SCC"
alias goto="<<S9900=\A"; fi
goto
uvlp13LD $NALIST #<-- send print file to laser printer <-- NOTE
#8======================================================================
S9000=A #<--Normal End of Job - see details on page '2A1'
S9900=A #<--Abnormal End of Job
NOTE 'uvlp13LD $NALIST' line inserted to send the print file to a laser printer. See the 'uvlp13LD' script listed at https://uvsoftware.ca/uvlist.htm#E3. This script (for the uvlist utility) contains laser printer codes to print at 13 cpi & 8 lpi, Landscape & Duplex. This allows you to print a mainframe report (132 columns wide & 66 lines deep) on 8 1/2 x 11 (Landscape & Duplex).
There are many other 'uvlp...' scripts for a variety of cpi, lpi, Landscape/Portrait, Duplex/Simplex, margins, etc. See uvlist.htm.
COBOL programs often create print reports via 'DISPLAY upon SYSOUT' and the JCL converter allows for this by inserting the following line on COBOL steps.
exportfile SYSOUT $SYOT/${jobid2}_${JSTEP}_SYSOUT_$(date +%y%m%d_%H%M%S)
========================================================================
SYSOUT was not used by our demo program (car100.cbl), but if it had been, you would print the report as follows:
uvlp13LD $SYSOUT #<-- send print file (SYSOUT vs NALIST) to laser printer #8==============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
You will often need to search the JCL/scripts to check conversion results. You can search for patterns that might confirm desired conversions or reveal problems that need to be corrected. 'grep' is the 1st choice if you only need to search for 1 pattern. The Vancouver Utilities has powerful tools to search for multiple patterns on 1 line or multiple lines in a file.
For example 'selectf1' will search entire directories for up to 5 patterns anywhere in each file for specified occurrence counts =, >, or < specified values. Default among multiple conditions is 'AND', but may specify 'OR'. See example to search for SORT's with 'INCLUDE COND' or 'OMIT COND'
| 4A1. | 'grep' example to search directory of JCLs for 'SORT's & converted |
| scripts for 'uvsort's. |
| 4B1. | 'selectf1' example to search JCLs for SORT's with 'INCLUDE COND' or |
| 'OMIT COND'. Could then search converted scripts for selects or deletes, | |
| but the JCL search gives you the filenames to check in the converted | |
| scripts directory. |
| Note |
|
| 4C1. | scan1d - uvcopy job to search all files in a directory for data lines with |
| - matches to 1 pattern optionally qualified by 2 other patterns. | |
| - might use prior to 'rep1d' (documented in Part_5). |
| 4D1. | scan2d - uvcopy job to search all files in a directory for data lines with |
| - matches to any 1 of multiple patterns in a pre-edited table file. | |
| - might use prior to 'rep2d' (documented in Part_5). | |
| that uses a table file of search/replace/qualify patterns. |
| 4E1. | scan3d - select lines from a starting pattern to an ending pattern, |
| optionally qualified by a 3rd pattern present anywhere between start & end. | |
| - see example selecting IKJEFT01 steps to see various programs being run |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'grep' is the 1st choice when searching directories for specified patterns, but you can only search for 1 pattern at a time. On the following pages, we will demonstrate 'selectf1' which will search for up to 5 patterns.
Page '4B3' will demonstrate searching the mvstest demo JCLs for SORTs, that have 'INCLUDE COND=...' or OMIT COND=...', but first we will show you the 'grep' method of searching for SORT's in JCLs. We can use grep to search for 'PGM=SORT' in the unconverted jcl2/* or for 'uvsort in the converted jcl3/* We will use the demo JCL's listed in JCLcnv1demo.htm#Part_2 & that you may have copied to your homedir/testlibs as instructed JCLcnv1demo.htm#Part_3.
/home/userxx <-- setup testlibs in your homedir :-----testlibs <-- $RUNLIBS libraries for JCL & COBOL conversions : :-----jcl0 - test/demo JCLs supplied : : :---jar100.jcl,jar200.jcl,jgl100.jcl,jgl200.jcl,jgl230.jcl,jgl320.jcl : :-----jcl1 - intermediate conversion 73-80 cleared : :-----jcl2 - JCLs with PROCs expanded from procs/... : :-----jcl3 - JCLs converted to Korn shell scripts
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#3. grep 'EXEC PGM=SORT' jcl2/*
===========================
jcl2/jar200.jcl:3://STEP010 EXEC PGM=SORT,REGION=2048K jcl2/jgl200.jcl:10://STEP010 EXEC PGM=SORT,REGION=2048K,PN=STEPA jcl2/jgl230.jcl:5://STEP010 EXEC PGM=SORT
So now we know which JCLs have SORTs, but we would then have to check each of them to see if they have 'INCLUDE COND=...' or 'OMIT COND=...'. Not a big problem when only 3 JCLs to check, but there could be hundreds of JCLs with SORTs, but only a few with the INCLUDEs/OMITs desired.
#4a. vi jcl2/jar200.jcl <-- see listing on page '2B1'
==================
#4b. vi jcl2/jgl200.jcl <-- see listing on page '2D1'
==================
#4c. vi jcl2/jgl230.jcl <-- see listing on page '2E1'
==================
See the page '4B3' for a better way using 'selectf1'.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'selectf1' will select filenames from a directory of files depending on search pattern occurrences = > < specified values. You may specify up to 3 search patterns with occurrence counts = or > or < than specified values. The default condition (among multiple search patterns) is 'AND', but you may specify 'OR'.
Here are a few examples (before presenting the command FORMAT & RULES below).
uvcopy "selectf1,fild1=cbl0,arg1u=EXEC_CICS:>0,arg2u=EXEC_SQL:>0" ================================================================= - select COBOL programs with 'EXEC CICS's AND 'EXEC SQL's
uvcopy "selectf1,fild1=cbl0,arg1u=EXEC_CICS:>0,arg2u=EXEC_SQL:=0" ================================================================= - select COBOL programs with 'EXEC CICS's AND WITHOUT (=0) 'EXEC SQL's
uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>1" ============================================= - select JCLs with more than 1 SORT
uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>0,arg2u=INCLUDE_COND:>0" =================================================================== - select JCLs with 1 or more SORTs 'AND' that have at least 1 INCLUDE CONDition
uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>0,arg2u=INCLUDE_COND:>0,arg3u=or:OMIT_COND:>0" ========================================================================================= - select JCLs with SORTs 'AND' with INCLUDE CONDitions 'OR' OMIT CONDitions
uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>0,arg2u=INCLUDE_COND:>0,arg3u=EXEC_PGM:<5" ===================================================================================== - select JCLs with 'SORT' AND 'INCLUDE COND' and with less than 5 steps
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
uvcopy "selectf1,fild1=indir,fild2=outdir,filo1=tmp/filenames,filo2=tmp/copyscript\
,arg1=pattern1:=><99,arg2=pattern2:=><99,arg3=or:pattern3:=<>99,arg4=...,arg5=..."
=========================================================================================
- command FORMAT
uvcopy selectf1,fild1=indir =========================== - minimum command line suggested - will be prompted for arg1,arg2,arg3,arg4,arg5 - enter null for arguments not required
| NOTE |
|
| fild1=... |
|
| fild2=... |
|
filo1=tmp/searchpattern1_list
- file of filenames selected
- defaults to searchpatter1 with '_list' appended
- will be prompted for command to display filenames at EOJ
- could enter vi, more, cat, etc
filo2=tmp/searchpattern1_copy
- script file of 'cp' commands to copy selected files to outdir
- defaults to searchpatter1 with '_copy' appended
- often ignored, when only interested in filenames
arg1=pattern1:=><99 - must specify pattern1
arg2=pattern1:=><99 - pattern2 optional arg2=or:pattern1:=><99 - may specify 'or:' on patterns 2,3,4,5 --- up to 5 search args --- arg5=pattern1:=><99 arg5=or:pattern1:=><99
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here we will present the same demo as for 'grep' on page '4A1', but using 'selectf1' which can search for up to 5 desired patterns with 1 command - here searching for 'SORT's with 'INCLUDE COND' OR 'OMIT COND'.
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#3. rm -f tmp/* tmp1/* <-- clear tmp/ dirs (used for output files)
#4. uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>0,arg2u=INCLUDE_COND:>0,arg3u=or:OMIT_COND:>0"
=========================================================================================
150330:201538:selectf1: EOF fili01 rds=21 size=929: jcl2/jar200.jcl
150330:201538:selectf1: EOF fili01 rds=16 size=674: jcl2/jgl100.jcl
150330:201538:selectf1: EOF fili01 rds=25 size=1123: jcl2/jgl200.jcl
150330:201538:selectf1: EOF fili01 rds=30 size=1301: jcl2/jgl230.jcl
150330:201538:selectf1: EOF fili01 rds=21 size=844: jcl2/jgl320.jcl
- - - 5 lines omitted - - -
150330:201538:selectf1: EOF fili02 rds=10 size=115: tmp/selectf1files
150330:201538:selectf1: EOF filo01 wrts=1 size=11: tmp/PGM_SORT_list 150330:201538:selectf1: EOF filo02 wrts=1 size=55: tmp/PGM_SORT_copy
#5. more tmp/PGM_SORT_list <-- display filenames selected
====================== - shown on 2 lines to save space
jar200.jcl <-- only 1 JCL has EXEC 'PGM=SORT' & 'OMIT COND=...'
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. Login uvadm --> /home/uvadm
#2. cd mvstest/testlibs
#3. uvcopy "selectf1,fild1=jcl0,arg1=PGM=SORT:>0,arg2u=INCLUDE_COND:>0,arg3u=or:OMIT_COND:>0"
=========================================================================================
150330:202746:selectf1: EOF fili01 rds=24 size=807: jcl2/adrdssu1.jcl
150330:202746:selectf1: EOF fili01 rds=8 size=354: jcl2/eztlist.jcl
150330:202746:selectf1: EOF fili01 rds=13 size=309: jcl2/ftpdemo1.jcl
150330:202746:selectf1: EOF fili01 rds=9 size=408: jcl2/jar100.jcl
150330:202746:selectf1: EOF fili01 rds=21 size=929: jcl2/jar200.jcl
- - - 83 lines omitted - - -
150330:202746:selectf1: EOF fili02 rds=88 size=1027: tmp/selectf1files
150330:202746:selectf1: EOF filo01 wrts=9 size=99: tmp/PGM_SORT_list 150330:202746:selectf1: EOF filo02 wrts=9 size=495: tmp/PGM_SORT_copy
#4. vi tmp/PGM_SORT_list <-- see filenames selected
==================== - 9 JCL's selected of 88 total
# selectf2 - select filenames from jcl2 - 2015/04/01_11:41:11
# EXEC_PGM=SORT:>0 INCLUDE_COND:>0 or:OMIT_COND:>0
jar200.jcl # 21
jar205.jcl # 21
jar410.jcl # 11
jar420.jcl # 12
jar430.jcl # 22
jar440.jcl # 11
jar450.jcl # 14
jgl210.jcl # 24
jgl215.jcl # 20
#5. vi tmp/PGM_SORT_list <-- see script to copy selected files to tmp1/...
==================== - ignore if all you need is list of filenames
cp jcl2/jar200.jcl tmp1
cp jcl2/jar205.jcl tmp1
--- 5 lines omitted --- tmp1
cp jcl2/jgl210.jcl tmp1
cp jcl2/jgl215.jcl tmp1
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Vancouver Utilities provides several general purpose 'scan/replace' utilities that can be more effective than 'grep' & 'sed' in complex situations because:
Suppose we wanted to scan a directory of mainframe JCL for all 'EXEC's with CONDition codes. Here are a few lines (of thousands):
//JOBA JOB .......... most lines not shown .......... //STEP020 EXEC PGM=IDCAMS,COND=(4,LT,STEP010)
//JOBB JOB .......... most lines not shown .......... //STEP050 EXEC PGM=CGL100,COND=(4,LT),REGION=1024K,PARM=$PARM1
uvcopy scan1d,fild1=jcl0,arg1=EXEC,arg2=PGM,arg3=COND,uop=p2 ============================================================ - could enter all info on the command line as above
uvcopy scan1d,fild1=jcl0,uop=p2 =============================== - OR you will be prompted for info not on the command line - see next page for console display prompts & entries
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
uvcopy scan1d,fild1=jcl0,uop=p2 ===============================
uop=q1a1b99999c0d256e0p1 - option defaults
a1 - begin scan at line #1 (default) of input file
a0 - causes prompt for search pattern to BEGIN scan
b99999 - end scan at line #99999 (dflt) of input file
b0 - causes prompt for search pattern to END scan
c0 - begin scan at col #1 (zero rel) of input lines
d256 - length of scan area default 256 bytes
e1 - inhibit errmsgs when LF not found within rcsz
p1 - 1 present qual(31-60) & 1 absent qual(61-80)
p2 - 2 present quals(31-60+61-80) & NO absent quals
p4 - 2 absent quals(31-60+61-80) & NO present quals
User OPtion (uop) defaults = q1a1b99999c0d256e0p1
null to accept or re-specify (1 or more) --> p2 <-- enter p2 2 present Quals
enter search pattern (max 64) --------------> EXEC <-- enter search pattern
enter 1st qualifier --------------> PGM <-- enter 1st qual
enter 2nd qualifier --------------> COND <-- entire 2nd qual
enter options: i,p,q1/q2/q3,n,or null ----->
080109:205346:scan1d: EOF filo01 wrts=47 size=2172: tmp/jcl0_EXEC hits=14
EOJ, Output File written to: tmp/jcl0_EXEC
enter command: vi,cat,more,lp,uvlp12,null -->
Job: scan1d Dir: jcl0 Search: EXEC Qual1: PGM Qual2: COND SearchBgn: SearchEnd: MatchOps: UserOps: q1a1b99999c0d256e0p1p2 ====================================================== 2008/01/09_20:53:46 00005 //STEP020 EXEC PGM=CGL200,COND=(4,LT,STEP010),PARM=&YEAREND 00010 1 hits @EOF: jcl0/JGL220.jcl
00011 //STEP020 EXEC PGM=IDCAMS,COND=(4,LT,STEP010) 00018 1 hits @EOF: jcl0/JGL320.jcl
00012 //STEP020 EXEC PGM=CAR200,COND=(4,LT,STEP010),PARM=2006 00017 1 hits @EOF: jcl0/JAR200.jcl
00006 //STEP020 EXEC PGM=IDCAMS,COND=(4,LT,STEP010) 00011 1 hits @EOF: jcl0/JPY402.jcl
====================================================== 2008/01/09_20:53:46 EOD: 00014 hits in 0014 files of 0046 total (00576 lines)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Suppose we wanted to find the EXEC's with COND= (similar to above scan1d), BUT only for certain utilties. We can edit a table of patterns for scan2d as follows. The table allows # comments.
#1. vi ctl/utilcond.tbl <-- prepare table of patterns
===================
# utilcond.tbl - table of search patterns for scan2d # - to find EXEC's with CONDs for a list of utilities # 01-30 - search pattern (ended by 1st tilde) # 31-60 - presence qualifier, must ALSO be present, disabled by tildes # 61-80 - absence qualifier, or 2nd presence qualifier if uop=p2 IDCAMS~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~~~ IEBGENER~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~~~ IKJEFT~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~~~ SORT~~~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
#2. uvcopy scan2d,fild1=jcl2,fili2=ctl/utilcond.tbl,uop=p2
======================================================
filo1=tmp/jcl0_utilcond.tbl <-- output file defaults to this
uop=p2 <--option p2 for 2 presence qualifiers (vs 1 present & 1 absent)
#3. vi tmp/jcl2_utilcond.tbl <-- investigate the outpout report
========================
Job: scan2d Dir: jcl2 MatchOps: ndg2 UserOps: q1a1b99999c0d256p1t1p2 TableFile: ctl/utilcond.tbl SearchBgn: SearchEnd: ====================================================== 2015/11/03_16:05:25 IDCAMS~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~ IEBGENER~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~ IKJEFT~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~ SORT~~~~~~~~~~~~~~~~~~~~~~~~~~EXEC~~~~~~~~~~~~~~~~~~~~~~~~~~COND=~~~~~~~~~~~~~ ====================================================== 2015/11/03_16:05:25 00010 //STEP020 EXEC PGM=IDCAMS,COND=(4,LT,STEP010) 00016 1 hits @EOF: jcl2/jgl300.jcl
00013 //STEP020 EXEC PGM=IEBGENER,COND=(4,LT,STEP010) 00021 1 hits @EOF: jcl2/jgl320.jcl
00013 //STEP020 EXEC PGM=IKJEFT01,COND=(4,LT,STEP010) 00027 1 hits @EOF: jcl2/jgl330.jcl
00013 //STEP020 EXEC PGM=IDCAMS,COND=(4,LT,STEP010) 00027 1 hits @EOF: jcl2/jgl331.jcl ====================================================== 2015/11/03_16:05:32 EOD: 00009 hits in 0009 files of 0090 total (01576 lines)
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'scan3d' will select lines from a starting pattern to an ending pattern, optionally qualified by a 3rd pattern present anywhere between start & end.
I have found this job very useful during JCL conversions, for example I wanted to investigate the various programs that were being run under IKJEFT01.
I will run scan3d on the converted scripts in jcl3/... (vs JCLs in jcl2/...) because I want to see the SYSIN & SYSTSIN parms that are inserted into the converted sripts.
IKJEFT01 steps will be something like:
##STEP0010 EXEC PGM=IKJEFT01 <-- start pattern PGM=IKJEFT
---- lines omitted ----
exportfile SYSTSIN $RUNLIBS/parms/ikjdemo1
---- lines omitted ----
#3----------------------------------------------------------------------
cobrun $ANIM $CBLX/car100
#4----------------------------------------------------------------------
<-- end pattern '#4---'
#1. uvcopy scan3d,fild1=jcl3,arg1=PGM=IKJEFT,arg2=#4---
===================================================
,filo1=tmp/$fild1_$arg1 <-- output file defaults to directory_pattern
#2. vi tmp/jcl3_PGMIKJEFT <-- inspect output report
=====================
Job=scan3d scan files in Dir: jcl3
Selecting From: "PGM=IKJEFT" To: "#4---" Qualified By: "~"
====================================================== 2015/11/03_16:27:13
00019 ##STEP0010 EXEC PGM=IKJEFT01,DYNAMNBR=20
00020 export PROGID=car100
00022 exportfile SYSTSIN $RUNLIBS/parms/ikjdemo1
00023 #2-------------------------- parms/ikjdemo1 ---------------------------
00024 # DSN SYSTEM(DB2P)
00025 # RUN PROGRAM(CAR100) PLAN(IKJPROG1) LIB('MEPP.WXS91.LOADLIB')
00026 # PARMS('parmsdata added on IKJEFT01 demo Aug28/07')
00027 # END
00030 #3----------------------------------------------------------------------
00031 cobrun $ANIM $CBLX/car100
00032 #4----------------------------------------------------------------------
00052 14 selected from 52 @EOF: jcl3/ikjeft01.ksh
====================================================== 2015/11/03_16:27:13
EOD: 1 selects, 14 lines in 1 files of 90 total (5906 lines)
The output report will includes all matches from all JCL/scripts in the directory and there could be multiple matches in some jobs.
If you want to output separate files for each match, you can use alternate job 'scan3dd' (vs 'scan3d') & specify ',fild2=outputdirectory'
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 5A1. | Mass changing all JCL/scripts - summary of tools available |
| 5B1. | mass change example using 'sed' & script 'runsed1' |
| - uncomment all '#rm' file deletes after test/debug | |
| - converter #comments out to assist test/debug | |
| - could remove #comment after test/debug (or could leave as is) | |
| - example using 'runsed1' to prompt for sed 's' substitute commands | |
| & apply while copying all files in directory to a 2nd directory |
| 5C1. | mass change example using sed, runsed1,& runsed2 |
| change all 'ksh' to 'ksh93', for clients that do not want | |
| to upgrade their current ksh (ksh88) to ksh93 (best for JCL/scripts) | |
| - uses 'runsed1' for the 1s of 4 directories to be processed | |
| - uses 'runsed2' to process other 3 using sed commands saved by runsed1 |
| 5D1. | mass change using uvcopy job 'rep1d' |
| rep1d - copies all files from 1 dir to a 2nd dir | |
| - prompts for 1 pattern to be replaced by a 2nd pattern | |
| - optionally qualified by 2 other patterns (present or absent) |
| 5E1. | mass change using uvcopy job 'rep2d' |
| rep2d - copies all files from 1 dir to a 2nd dir | |
| - searching/replacing depending on a pre-edited table | |
| of search/replace/qualify patterns |
| 5F1. | rep3d - replace specified blocks of code in all files in a directory |
| - blocks identified by a start pattern, an end pattern,& | |
| an optional qualifier pattern between start & end patterns | |
| - replacement block must be pre-edited into a text file |
| 5G1. | uvfix1 - easy way to fix various problems in datafiles & scripts. |
| - allows you to apply the powerful uvcopy instructions | |
| - translates, convert data field type (packed/binary/zoned) | |
| - check output in tmp/... if OK - copy back to input directory | |
| - example given to correct EBCDIC zoned signs to unix standards |
| 5H1. | uvfixA - same as uvfix1, but processes all files in a directory |
| while copying to the tmp1/ directory | |
| - check output & if OK - copy back to input directory |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| sed |
|
| runsed1 |
|
| runsed2 |
|
| rep1d |
|
| rep2d |
|
| scan1d |
|
| scan2d |
|
| uvfix1 |
|
| uvfixA |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The JCL converter options control file may be coded with option 'v3', which causes all DELETEs to be #commnented when converted to unix 'rm's (#rm). This will help you debug jobs since the intermediate files often give you clues to the problem. Note that unix does not have a problem overwriting files as did the mainframe.
After test/debug complete you could run a sed script to mass change all '#rm's to 'rm's. I suggest using 'runsed1' (script in $UV/sf/util).
#1. Login userxx --> /home/userxx #2. cdl --> /home/userxx/testlibs (alias cdl='cd $RUNLIBS')
#3. mv jcls jcls.old <-- save (change name) of existing jcls subdir
================
#4. mkdir jcls <-- remake new dir
==========
#5. runsed1 jcls.old jcls <-- run script (will copy files to new jcls)
=====================
--> s/^#rm/rm/ <-- enter sed command at prompt
========== '^' to match at begining of line
--> control-D <-- control-D to end command entry
#6. diff jcls.old/jobxx.ksh jcls/jobxx.ksh <-- spot check results ?
======================================
-- OR use the 'alldiff2' script to check the entire directory
#5. alldiff2 jcls.old jcls <-- creates alldiff report in tmp/jcls.dif
======================
#5a. vi tmp/jcls.dif <-- examine alldiff report
===============
'alldiff2' is great after mass changes, to ensure yu changed what you intended and with no unintended changes.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Here is an example of a mass change that may be required at some sites that are not running a current version of the Korn shell (required for Vancouver Utilities & JCL conversions to Korn shell scripts).
Download (free) current version of ksh (1993) from www.kornshell.com Look for ksh standalone AT&T executable for your machine (Linux,AIX,SUN,HP). You do not need the source version, just download & copy into /bin because the 1st line of all VU scripts (& converted JCL/scripts) is '#!/bin/ksh'.
But what do you do if the customer does not want to update to the latest version & does not want you to modify any system directories such as /bin ?
We can download the lastest version (1993) of ksh, change the name to ksh93, and store it in /home/uvadm/bin/ksh93 (vs /bin/ksh). We then need to change line 1 of all VU scripts & modify ctl/jclunixop51 to convert JCLs to scripts with line 1 coded as '#!/home/uvadm/bin/ksh93' (vs '#!/bin/ksh').
#1. Internet browser to www.kornshell.com
#2. Look for ksh standalone AT&T executable for your machine (Linux,AIX,SUN,HP).
#3. Download into /home/uvadm/bin/ksh
#4. mv /home/uvadm/bin/ksh /home/uvadm/bin/ksh93 <-- rename as 'ksh93'
============================================
If you have root permissions, you can simplify things by storing ksh93 in /bin/ksh93 vs /home/uvadm/bin/ksh93.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Vancouver Utility scripts are split into 4 subdirs as shown below. We will save current subdirs as adm.old, demo.old, IBM.old, util.old, make new subdirs & run mass change script 'runsed1' to modify.
/home/uvadm :-----sf : :-----adm : :-----demo : :-----IBM : :-----util
#1. Login uvadm --> /home/uvadm
#2. cd sf
#3. mv adm adm.old; mv demo demo.old; mv IBM IBM.old; mv util util.old
==================================================================
- save existing subdirs by renaming to ---.old
#4. mkdir adm demo IBM util tmp
===========================
- make new subdirs to receive modified scripts
#5. util.old/runsed1 adm.old adm
============================
- runsed1 prompts for sed command, enter as follows:
#5a. --> s:#!/bin/ksh:#!/home/uvadm/bin/ksh93: <-- enter sed substitute command
=====================================
#5b. --> ^D <-- enter control-D to end command entry
| Note |
|
#6a. util.old/runsed2 demo.old demo tmp/runsed1
==========================================
#6b. util.old/runsed2 IBM.old IBM tmp/runsed1
==========================================
#6c. util.old/runsed2 util.old util tmp/runsed1
==========================================
#7a. chmod 775 adm/*; chmod 775 demo/*; chmod 775 IBM/*; chmod 775 util/*
====================================================================
- change permissions back to 775 (adding execute bits)
#8. Optional - could run alldiff2 to verify changes as intended
#8a. alldiff2 adm.old adm <-- create tmp/adm.dif (check with vi)
====================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The JCL converter generates 1st line of converted JCL/scripts as '#!/bin/ksh'. We must now change the control file to generate as '#!/bin/ksh93'. or '#!/home/uvadm/bin/ksh93' if not allowed to modify /bin or /usr/bin. Here are the 1st 2 lines & lines 86-101.
jclunixop51:a2b3c0d1e2f0g1i15j0k15l50m4n3o0p0r0s0t1u1v3w0x0y1z1 #<-- demo options
# ===================================================
--- lines 3-85 omitted ---
# Following lines will be inserted in scripts until :ENDINSERT
# - except for '# ' comment (col1 '#', col2 ' ' blank)
# '#!/.......' is 1st line output - usually '#!/bin/ksh' or '#!/bin/ksh93'
#!/bin/ksh
export JOBID2=JOBIDX jobid2=jobidx; scriptpath="$0"; args="$*"
if [[ -z "$JOBID1" ]]; then export JOBID1=$JOBID2; fi
for arg in $args; do if [[ "$arg" == *=* ]]; then export $arg; fi; done
integer JCC=0 SCC=0 LCC=0 # init step status return codes
autoload jobset51 gdgclear51 jobend51 jobabend51 logmsg1 logmsg2 stepctl51
autoload exportfile exportgen0 exportgen1 exportgen2 exportgenall exportgenx
. $APPSADM/env/stub.ini #<-- for control-M (see notes in env/stub.ini)
jobset51 # call function for JCL/script initialization
#gdgclear51 $PUB_... #<-- insert gdgclear51's for exportgen1's to $PUBs
goto
S0000=A
:ENDINSERT: - above non #comment lines inserted at begin output scripts
#1. Login uvadm -->
#2. vi ctl/jclunixop51 <-- edit JCL converter control file
==================
#2a. --> #!/home/uvadm/bin/ksh93 <-- change line 89 (from #!/bin/ksh)
=======================
#3a. cp /ctl/jclunixop51 /home/userxx/testlibs/ctl
=============================================
#4a. Login userxx --> /home/userxx (or whatever Login desired)
#4b. cdl --> alias cdl='cd $RUNLIBS' --> /home/userxx/testlibs
#4c. vi ctl/jclunixop51
==================
#4d. --> #!/home/uvadm/bin/ksh93 <-- change line 89 to '#!/home/uvadm/bin/ksh'
=======================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'rep1d' is a Vancouver Utility pre-programmed uvcopy job to copy all files from 1 directory to another, while changing any 1 specified pattern to an alternate. 'rep1d' provides qualifiers & various options.
Suppose we wanted to insert 'acme.' prior to all filenames that begin with 'ar.' in the JCL/scripts. Here is an example of the BEFORE & AFTER desired.
exportfile CUSTMAS data1/ar.customer.master <-- BEFORE change ================================================ exportfile CUSTMAS data1/acme.ar.customer.master <-- AFTER change
We recommend you first change the name of the existing 'jcls' to 'jcls.old', make a new empty 'jcls' subdir, run the mass change while copying files from jcls.old to jcls,& then verify the results using the 'alldiff' script.
#1. mv jcls jcls.old - rename existing subdir (for backup)
================
#2. mkdir jcls - make new empty subdir (to receive changes)
==========
#3. uvcopy rep1d,fild1=jcls.old,fild2=jcls
======================================
uop=q1a1b99999c0d256e0p1 - option defaults
a1 - begin scan at line #1 (default) of input file
a0 - causes prompt for search pattern to BEGIN scan
b99999 - end scan at line #99999 (dflt) of input file
b0 - causes prompt for search pattern to END scan
c0 - begin scan at col #1 (zero rel) of input lines
d256 - length of scan area default 256 bytes
e0 - do not erase any output files
e1 - erase output files with no replacements
p1 - 1 present qual(31-60) & 1 absence qual(61-80)
p2 - 2 present quals(31-60+61-80) & NO absent quals
User OPtion (uop) defaults = q1a1b99999c0d256e0p1
null to accept or re-specify (1 or more) -----> <-- defaults OK
did you create tmp outdir or remove files -----> y <--
enter your search pattern (max 64 bytes) ------> ar. <-- SEARCH pattern
enter replacement pattern (max 64 bytes) ------> acme.ar. <-- REPLACE pattern
enter qualifier present (null to disable) -----> <-- n/r
enter absent qual (2nd present if optn p2) ----> <-- n/r
match options: p(@,#,etc), s(in replace)
quote inhibits: q1=singles,q2=doubles,q3=both
enter match options: p,s,q1/q2/q3,n,or null ---> <-- n/r
EOJ, OutFile written to: tmp/jcls.old_tgl0
enter: vi,cat,more,lp,uvlp12,etc (or null) ----> more <-- display audit file
Please see the 'rep1d' audit file on the next page --->
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'rep1d' creates an audit file of the 'after changes' images of any lines modified by the search/replace patterns. By default the audit file is created in the 'tmp' subdir & the filename is a concatenation of the input directory & the search pattern. For example: tmp/jcls_ar.
Job: rep1d InDir: jcls.old OutDir: jcls Search: ar. Replace:acme.ar. QualPresent: export Absent: SearchBgn: SearchEnd: MatchOptions: UserOptions: q1a1b99999c0d256e0p1 ====================================================== 2015/01/25_15:48:08 00023 exportfile CUSTMAS data1/acme.ar.customer.master 00024 exportfile NALIST data1/acme.ar.customer.nameadrs.list100 00044 2 reps @EOF: jcls/jar100.ksh 00023 exportfile SORTIN data1/acme.ar.sales.items 00055 exportfile CUSTMAS data1/acme.ar.customer.master.indexed 00056 exportfile SALELST data1/acme.ar.sales.list 00075 3 reps @EOF: jcls/jar200.ksh ====================================================== 2015/01/25_15:48:08 EOD: 5 reps in 2 files of 9 total (582 lines)
#4. alldiff jcls.old jcls
=====================
23,24c23,24 < exportfile CUSTMAS data1/ar.customer.master < exportfile NALIST data1/ar.customer.nameadrs.list100 --- > exportfile CUSTMAS data1/acme.ar.customer.master > exportfile NALIST data1/acme.ar.customer.nameadrs.list100 diff file# 2 - jcls.old/... vs jcls/jar100.ksh
23c23 < exportfile SORTIN data1/ar.sales.items --- > exportfile SORTIN data1/acme.ar.sales.items 55,56c55,56 < exportfile CUSTMAS data1/ar.customer.master.indexed < exportfile SALELST data1/ar.sales.list --- > exportfile CUSTMAS data1/acme.ar.customer.master.indexed > exportfile SALELST data1/acme.ar.sales.list diff file# 3 - jcls.old/... vs jcls/jar200.ksh
2 different of 9 files compared jcls.old to jcls
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'rep2d' copies all files from 1 dir to a 2nd, while performing mass change via a pre-edited table of search/replace patterns. 'rep1d' is handy when you only have 1 search/replace pair (because it solicits the patterns), but 'rep2d' is more powerful (because it allows an unlimited number of search/replace patterns in the pre-edited table file).
#1. vi ctl/rep2d.tbl <-- create table file of search/replace patterns
================ (in this case, inspect supplied demo table)
# rep2d.tbl - search/replace table to demo uvcopy rep2d # - insert acme. prefix on filenames ar... & gl... # 01-30 search 31-60 replace 61-80 qualify ar.~~~~~~~~~~~~~~~~~~~~~~~~~~~acme.ar.~~~~~~~~~~~~~~~~~~~~~~export~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ gl.~~~~~~~~~~~~~~~~~~~~~~~~~~~acme.gl.~~~~~~~~~~~~~~~~~~~~~~export~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
#2a. mv jcls jcls.old <-- change name of existing jcls (to backup file)
================
#2b. mkdir jcls <-- create new subdir to receive updated files
==========
#3. uvcopy rep2d,fild1=jcls.old,fild2=jcls,fili3=ctl/rep2d.tbl
==========================================================
- copy jcls.old to jcls with search/replace
- also creates an audit file (tmp/jcls.old_rep2d.tbl)
uop=q1a1b99999c0d256e0p1 - option defaults
a1 - begin scan at line #1 (default) of input file
a0 - causes prompt for search pattern to BEGIN scan
b99999 - end scan at line #99999 (dflt) of input file
b0 - causes prompt for search pattern to END scan
c0 - begin scan at col #1 (zero rel) of input lines
d256 - length of scan area default 256 bytes
e0 - do not erase any output files
e1 - erase output files with no replacements
p1 - 1 present qual(31-60) & 1 absent qual(61-80)
p2 - 2 present quals(31-60+61-80) & NO absent quals
User OPtion (uop) defaults = q1a1b99999c0d256e0p1
null to accept or re-specify (1 or more) ------> <-- defaults OK
did you create outdir or remove all files ------> y <--
match options: p=matchchars(@,#), s=also replace
quote inhibits: q1=singles,q2=doubles,q3=both
enter match options: p,s,q1/q2/q3,n,or null ----> <-- n/r
EOJ, OutFile written to: tmp/jcls.old_rep2d.demo
enter command: vi,cat,more,uvlp12,etc(or null)--> more <-- display audit rpt
Please see audit file report on the next page --->
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Job: rep2d InDir: jcls.old OutDir: jcls Table: ctl/rep2d.tbl SearchBgn: SearchEnd: MatchOptns: df2g2 UserOptns: q1a1b99999c0d256e0p1 ====================================================== 2015/01/25_15:49:27 ar.acme.ar.export gl.acme.gl.export ====================================================== 2015/01/25_15:49:27 00023 exportfile CUSTMAS data1/acme.ar.customer.master 00024 exportfile NALIST data1/acme.ar.customer.nameadrs.list100 00044 2 reps @EOF: jcls/jar100.ksh 00023 exportfile SORTIN data1/acme.ar.sales.items 00055 exportfile CUSTMAS data1/acme.ar.customer.master.indexed 00056 exportfile SALELST data1/acme.ar.sales.list 00075 3 reps @EOF: jcls/jar200.ksh 00028 exportgen0 0 ACCTMAS data1/acme.gl.account.master_ 00029 exportgen1 +1 ACTLIST data1/acme.gl.account.acntlist_ 00051 2 reps @EOF: jcls/jgl100.ksh 00029 exportfile SORTIN data1/acme.gl.account.tran1 00030 exportgen1 +1 SORTOUT data1/acme.gl.account.trans_ 00057 exportgen1 +1 GLTRANS data1/acme.gl.account.trans_ 00059 exportgen0 0 GLMSOLD data1/acme.gl.account.master_ 00061 exportgen1 +1 GLMSNEW data1/acme.gl.account.master_ 00080 5 reps @EOF: jcls/jgl200.ksh 00024 exportfile SORTIN data1/acme.gl.account.tran1 00025 exportgen1 +1 SORTOUT data1/acme.gl.account.trans_ 00051 exportgen1 +1 GLTRANS data1/acme.gl.account.trans_ 00053 exportgen0 0 GLMSOLD data1/acme.gl.account.master_ 00054 exportgen1 +1 GLMSNEW data1/acme.gl.account.master_ 00074 exportgen0 0 ACCTMAS data1/acme.gl.account.master_ 00075 exportgen1 +1 ACTLIST data1/acme.gl.account.acntlist_ #(MOD,KEEP,DELETE) 00097 7 reps @EOF: jcls/jgl230.ksh 00026 exportgen0 0 SYSUT1 data1/acme.gl.account.master_ 00051 exportgen1 +1 ACNTMAS data1/acme.gl.account.master_ 00077 2 reps @EOF: jcls/jgl320.ksh ====================================================== 2015/01/25_15:49:31 EOD: 00024 reps in 0008 files of 0009 total (00582 lines)
#4. alldiff jcls.old jcls >tmp/jcls.dif <-- create alldiff report
===================================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'rep3d' will replace specified blocks of code in all files in a directory. Blocks of code are identified by a start pattern, an end pattern,& an optional qualifier pattern between start & end patterns. The replacement block must be pre-edited into a text file.
For example, let's assume we want to replace the FTP commands as follows:
open 214.69.21.9 <-- existing IP# (& start pattern)
mvstest mvstest11
cd ftpdata1
binary
open $IP_214_69_21_9 <-- replace with $variable
mvstest mvstest22 <-- change password
pwd <-- insert pwd for console log
cd ftpdata2 <-- change directories
binary <-- end pattern for block of code
#0a. login
=====
#0b. cdl (alias cd $RUNLIBS) --> /home/userxx/testlibs/
#1. vi ctl/ftp2 <-- edit the replacement block as shown above
===========
#2. mkdir tmp tmp1 <-- make output subdirs
==============
#2a. rm -f tmp/* tmp1/* - or remove all files if tmp1 already exists
#3. uvcopy rep3d,fild1=jcl3,fild2=tmp1,fili3=ctl/ftp2,filo4=tmp/rep3d.audit
=======================================================================
prompt begin pattern--> open 214.69.21.9
prompt end pattern ---> binary
prompt qualifier -----> =mvstest '=' present or '!' not present
prompt match options--> <-- null reply (or 'i' case insensetive)
#4. vi tmp/rep2d.audit <-- check audit report replacements made
==================
#5. vi tmp1/* <-- check output JCL/scripts
=========
#6. mv jcl3 jcl3.old <-- save original
================
#7. mv tmp1 jcl3 <-- rename tmp1 to jcl3 if AOK
============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'uvfix1' is a script that makes it easy to use the powerful repertoire of uvcopy instructions to fix various data file problems. See all uvcopy instructions documented at https://uvsoftware.ca/uvcopy3.htm.
The 'uvfix1' script saves having to write the framework of file I/O and prompts you for the instructions to fix your data problem. For example, suppose you have FTP'd a GL transaction file in text mode which translates to ASCII, but any EBCDIC negative zone signs will appear as shown below in our gltranE demo file. Negatives '0123456789' apear as '}JKLMNOPQRS'.
1 2 3 4 5
12345678901234567890123456789012345678901234567890
00011100T BANK FEE 000000100}
00061500T BACKUP TAPE DRIVE 000011508J
00061600T SONY BACKUP TAPES 000000805K
00062200T STAPLES OFFICE SUPPLIES 000001152L
00075300T TELEPHONE 000000525M
00083700T HYDRO 000000034N
00083800T WATER 000000109O
00073700T FEDEX 000000109P
00075100T OFFICE RENT 000003000Q
00063300T DOLLAR RENT A CAR 000001417R
For unix, negative signs for -0123456789 must be translated to 'pqrstuvwxy', because the bit pattern for these letters is the same as numerics 0123456789 with a 4 bit added, ie 0=x'30' & p=x'70', 1=x'31' & q=x'71', etc.
uvcopy provides the 'trt' instruction & translate table '$trtsea' to convert EBCDIC the sign convention to the unix sign convention used by Micro Focus & AIX COBOL & Vancouver Utilities.
trt b49(1),$trtsea <-- uvcopy instruction to convert EBCDIC signs to unix ==================
Normally you need to write a complete 'uvcopy job' with a framework of I/O file definitions & get/put instructions - as follows:
# fixsigns - uvcopy job to convert EBCIC sign convention to unix
# - for files FTP'd in text/ASCII mode
fili1=data1/gltranE,rcs=80,typ=LST
filo1=tmp/gltranA,rcs=80,typ=LSTt
@run
opn all open files
loop get fili1,a0(80) get next record into area 'a'
skp> eof (cc set > at EOF)
mvc b0(80),a0 move inarea 'a' to outarea 'b'
trt b49(1),$trtsea translate Sign in col 50 (49 zero relative)
put filo1,b0(80) write record to output fiel
skp loop return to get next record
eof cls all EOF - close files
eoj - end job
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. Login userxx --> /home/userxx (or whatever Login desired) #2. cdd --> /home/userxx/testdata (alias cdd='cd $RUNDATA')
#3. uvfix1 data1/gltranE <-- uvfix1 prompts as follows:
====================
uvfix1 - copy a text file & modify via uvcopy instruction keyins
- input rec in area a has already been copied to outputrec in b
uop=q1n99r8192 - user option defaults
n99 - output record count limit (high# copies to EOF)
r8192 - recsize ONLY for Fixed-Length OUTPUT records if desired
- for Fixed-Length, reply RSF/RST at prompt (default text typ=LSTt)
- for Text files: accept default recsize & default typ=LSTt
User OPtion (uop) defaults = q1r8192n999999999
-->null to accept or enter/override --> <-- null entry for text files
LSTt = typ default - accept or new typ --> <-- null entry accept LSTt
enter a uvcopy instruction (period "." terminates entries)
entry --> trt b49(1),$trtsea <-- entry
==================
entry --> . <-- enter '.' to terminate
150125:111159:uvfix1: EOF fili01 rds=10 size=510: data1/gltranE 150125:111159:uvfix1: EOF filo01 wrts=10 size=510: tmp/gltranE
cat tmp/gltranE <-- inspect output to confirm desired results =============== 00011100T BANK FEE 000000100p 00061500T BACKUP TAPE DRIVE 000011508q 00061600T SONY BACKUP TAPES 000000805r 00062200T STAPLES OFFICE SUPPLIES 000001152s 00075300T TELEPHONE 000000525t 00083700T HYDRO 000000034u 00083800T WATER 000000109v 00073700T FEDEX 000000109w 00075100T OFFICE RENT 000003000x 00063300T DOLLAR RENT A CAR 000001417y
cp tmp/gltranE data1/gltranA ============================ - copy back to data1/... changing name to preserve original for this demo - normally you would overwrite same name
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 6A1. | add data file info onto JCL/script file definitions (as #comments) |
| - appends Record-Sizes & Packed/Binary | |
| - examples of before/after adding datafileinfo as #comments |
| 6A2. | ctl/datactl53I - control file info for JCL conversion |
| 6B1. | add data file info onto JCL/script file definitions (as #comments) |
| - step by step instructions |
| 6C1. | cblcode1 - determine COBOL program type |
| - reads a directory of COBOL programs & writes a file of filenames | |
| with extensions to identify the program type. | |
| - filenames wwith suffix .cbl, .cdb, .cic depending presence of | |
| EXEC SQL, EXEC CICS, or neither. | |
| - also writes a 2nd file script to copy the directory to a 2nd | |
| directory appending the .cbl/.cdb/.cic extensions |
Might add 'IP# replace' to part 6 here in JCLcnv3aids.doc from JCL2cnvreal.doc#1E5 - 1E7 & 2A1
Might add jclunixop51 listing somewhere in this doc ?
Might add chkrcs1, chkrcs2, chkrcs2A
uvcopy chkrcs1,fild1=jcl3,filo2=rpts/chkrcs1.rpt ================================================ - check record sizes on uvsort/uvcp in JCL/scripts - see MVSJCL.doc or www.uvsoftware.ca/mvsjcl.htm
uvcopy chkrcs2,fild1=jcl3,filo1=rpts/chkrcs2_fj,filo2=rpts/chkrcs2_jf ===================================================================== - check record sizes on uvsort/uvcp in JCL/scripts - report datafilenames for rcs=99999 from saved SORTIN exports
chkrcs2A jcl3 <-- execute uvcopy job chkrcs2 + sorts =============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The converted JCL/scripts are in jcl3/..., but need to be copied to jcls/... for execution since only jcls is in $PATH (defined in the common_profile).
In November 2013, we added another step in the JCL conversion - running script 'jcldatafileinfo' which copies jcl3/* to jcl4/... appending COBOL I/O & record-size info (as #comments) on the right side of data file definitions, for example, here are the BEFORE & AFTER data files for step2 of jgl230.jcl. See complete listing beginning on page '2E1'.
exportgen1 +1 GLTRANS data1/gl.account.trans_
exportgen0 0 GLMSOLD data1/gl.account.master_
exportgen1 +1 GLMSNEW data1/gl.account.master_
exportgen1 +1 GLTRANS data1/gl.account.trans_ #cblIO=I/SS,rsz=80
exportgen0 0 GLMSOLD data1/gl.account.master_ #cblIO=I/SS,rsz=128p
exportgen1 +1 GLMSNEW data1/gl.account.master_ #cblIO=O/SS,rsz=128p
The appended #comments provide INVALUABLE info for JCL/script test/debug:
Note that 'cobfiles5A' is a Pre-Requisite for jcldatafileinfo. It creates Indexed file ctl/cobfilesI.dat/.idx with file info extracted from the COBOL programs & copybooks.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
ar.customer.master cpy=____________ rcs=00256 typ=RSF data=p job=jar100 prg=CAR100 ar.sales.items cpy=____________ rcs=00064 typ=LST data= job=jar200 prg=SORT gl.account.acntlist_ cpy=____________ rcs=00133 typ=LST data= job=jgl100 prg=CGL100 gl.account.master_ cpy=____________ rcs=00128 typ=RSF data=p job=jgl230 prg=CGL200 gl.account.trans_ cpy=____________ rcs=00080 typ=LST data= job=jgl230 prg=SORT gl.parms@pgl200s1 cpy=____________ rcs=_____ typ=RSF data= job=jgl200 prg=SORT py.payroll.master cpy=____________ rcs=00080 typ=RSF data=pb job=jpy200 prg=CPY200
'jcl2ksh51A' (run on page '3F1') created ctl/datactl53I, an Indexed control file with data file info (record-sizes, file-types,etc). This file is updated by cobfiles5A, mvsfiles5A,& jcldatafileinfo - so you can improve the conversions by running these jobs & re-converting with 'jclxx51' to get more record-sizes & file types coded on the JCL/scripts.
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#1. Login userxx --> /home/userxx (or whatever Login desired)
#2. cdl <-- alias cdl='cd $RUNLIBS' --> cd testlibs
#3. cobfiles5A cbls cpys maps
=========================
- create Indexed file info for jcldatafileinfo
#3a. uvlp12 ctl/cobfiles <-- good idea to print
=================== - useful report to assist test/debug
#4. mvsfiles5A jcl2 <-- updates ctl/datactl53I with rec-sizes & packed indicators
===============
#4a. uvlp12D xmvs/mvsfiles5 <-- could print file info report,
====================== - but too big for all JCL, better to print
for 1 JCL (after mvsfiles51 vs mvsfiles5A)
| Note |
#5. jcldatafileinfo jcl3 jcl4
=========================
- copy jcl3/* to jcl4, appending #cmts of COBOL I/O, recsize,Packed info
- invaluable during test/debug
- also updates ctl/datactl53I to improve conversion when re-converted
- prompts for option f1 ('typ=LST' on uvsort/uvcp files w/o packed/Binary)
---> f0 <--- reply f0 if you have many files with Packed/Binary
---> f1 <--- reply f1 if you have very few files with P/B
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#6. jclxx51 jcl2 jcl3 <-- reconvert all JCL jcl2/* --> jcl3/...
================= - after cobfiles5A, mvsfiles5A,& jcldatafileinfo
- update ctl/datactl53I record-sizes & packed indics
#7. jcldatafileinfo jcl3 jcl4 <-- rerun to #encode COBOL I/O & rec-sizes & p indics
=========================
#8. cp jcl4/* jcls <-- recopy to jcls/... (directory in $PATH)
==============
- would not copy all for real conversion if you have edited a lot of scripts
- OK here for test/demos
#8a. cp jcl4/jobnamex.ksh jcls <-- copy 1 script at a time before test/debug
========================= - recommended for your "real conversion"
You should NOT copy all jcl4/* jcls/ for your "real conversion" project. For your real conversion project, you should copy 1 script at a time when you ready to start test/debug on each script. This is an easy way to keep track of your progress (jcls/... are debugged, jcl4/... not yet)
Another reason NOT to copy all jcl4/* jcls/ when JCL/scripts re-converted:
If you are re-converting after changing options, it is a good idea to 1st save the current conversion results in jcl4/ as jcl4.old, reconvert,& then run script 'alldiff2' which creates a report (tmp/jcl4.dif) showing the differences in the scripts (caused by option changes).
#1. mv jcl4 jcl4.old <-- save current results as .old #2. mkdir jcl4 <-- make new dir for re-conversions #3. vi ctl/jclunixop51 <-- change options as desired
#6. jclxx51 jcl2 jcl3 <-- reconvert (same as above) #7. jcldatafileinfo jcl3 jcl4 <-- rerun #encode COBOL I/O, rec-sizes & p/b indics
#8. alldiff2 jcl4.old jcl4 <-- create differences report
======================
#8a. vi tmp/jcl4.dif <-- investigate differences
===============
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'jcl2ksh51A' (run on page '3F1') created ctl/datactl53I, an Indexed control file with data file info (record-size, etc). This file is updated by jcldatafileinfo and it might help to re-convert with jclxx51 to get more record-sizes coded on the JCL/scripts. If you wanted to update manually, you should unload the Indexed file to sequential file ctl/datactl53, edit,& reload the Indexed file as follows:
#1. uvcp "fili1=ctl/datactl53I,rcs=191,typ=ISF,filo1=ctl/datactl53,typ=LSTt"
========================================================================
- convert Indexed file back to sequential text file for 'vi' editor changes
#2. vi ctl/datactl53 <-- edit data control file
================ - update record-sizes, file-types
#3. uvcp "fili1=ctl/datactl53,rcs=191,typ=LST,key1=0(44),filo1=ctl/datactl53I\
,typ=ISF,isk1=0(44),trl=0(44)"
==========================================================================
- reload Indexed file for JCL converter
--- OR (easier to key) ---
#3a. uvcopy gdgload51 <-- reload Indexed file (alternate/easier method)
================
#4. jclxx51 jcl2 jcl3 <-- reconvert all JCL jcl2/* --> jcl3/...
================= - after option changes
#5. jcldatafileinfo jcl3 jcl4 <-- rerun to #encode COBOL I/O & rec-sizes
=========================
#6. cp jcl4/* jcls <-- recopy to jcls/... (directory in $PATH)
==============
- would not copy all for real conversion if you have edited a lot of scripts
- OK here for test/demos
Also note scripts 'mvsfiles51/mvsfiles5A' to determine the essential INPUT files required to run a job (ignoring intermediate files (outputs for later inputs). The most useful report is 'xmvs/mvsfiles5' which '*' flags the INPUT files actually needed to run any 1 job or all jobs. This report eliminates duplicates & intermediate files & identifies the essential files required to be be transfered from the mainframe. See details on pages '1L1' - '1L8'.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'cblcode1' is a uvcopy job to code COBOL source programs to 3 types by appending suffix .cbl, .cdb, .cic depending presence of EXEC SQL, EXEC CICS, or neither. It reads all programs in the directory, determines the type and writes 2 files:
#1. A file of filenames coded with .cbl/.cdb/.cic
#2. A script to copy the input directory of COBOL programs to a 2nd directory
appending .cbl/.cdb/.cic extensions, depending on presence of 'EXEC CICS'
or 'EXEC SQL' or neither
uvcopy cblcode1,fild1=cbl0,filo1=ctl/cbl0coded,filo2=sf/cbl0copycode ===================================================================== - here is a sample of the 2 output files:
# cblcode1 - 1st output ctl/cbl0coded filenames with ext .cbl/.cdb/.cic a04621a.cbl dehp0405.cdb ytrans1.cbl
# cblcode1 - 2nd output sf/cbl0copycode script to copy to cbl0c/filename.cbl/.cdb..cic cp cbl0/a04621a.cbl cbl0c/a04621a.cbl cp cbl0/dehp0405.cbl cbl0c/dehp0405.cdb cp cbl0/ytrans1.cbl cbl0c/ytrans1.cbl
'cblcode1' is executed by 'xrefall' & output file of filenames coded with .cbl .cdb .cic used to flag program types on crossrefs no flag on .cbl, '+' flag on EXEC SQL, '@' flag on EXEC CICS
The 2nd output file (script to copy to 2nd directory appending extensions) is not used by xrefall, but you might have occasion to use it. If you wanted to split the 3 types of COBOL programs to 3 separate directories you could run the ctl/cbl0coded script & then simply move extensions .cdb & .cic to seprate subdirs
#1. ctl/cbl0coded - script copies cbl0/... to cbl0c/... coding .cbl/.cdb/.cic
=============
#2. mv cbl0c/*.cdb cbl0cdb <-- move .cdb programs to separate directory
======================
#3. mv cbl0c/*.cdb cbl0cic <-- move .cic programs to separate directory
======================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
| 7A1. | Review of job-logging & console-logging |
| 7B1. | Test 'joblog2' run JCL/scripts serially - prove it works using jar100.ksh |
| - insert prompt/reply, run job from 2 screens, note delay | |
| - answer prompt on 1st job to complete & allow 2nd job to run |
| 7C1. | gdgreset1 - test automatic GDG generation# re-sequence at 900000 |
| 'gdgrest1' resets gdg#s for 'ONE' file automatically, whenever gdg# | |
| 900000 exceeded. 'gdgreset1' is called by function 'jobend51' which was | |
| inserted by the JCL converter at the end of all JCL/scripts. |
| 7D1. | gdgreset2 - test batch job to re-sequence gdg#s > specified threshold |
| 'gdgreset2' will reset GDG#s for 'ALL' files whose GDG#s exceed a | |
| specified threshold (suggested default 500000). You would run this | |
| batch job when no JCL/scripts are being executed. |
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Console-Logging captures the entire login session from login to logout (vs the 'joblog' script which captures the log only for 1 job at a time).
'console-logging' is better for batch shifts where an operator is manually running the JCL/scripts, 'better' because it captures the entire login session, including any commands between jobs and any responses to prompts.
'job-logging' is better for programmers who are running 1 job at a time and may want to review the log for jobs that roll off the screen. See 3 joblog scripts reviewed on the next page.
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
Three versions of the joblog script are provided:
| joblog |
|
| joblog1 |
|
| joblog2 |
|
joblog jar100.ksh <-- run jar100.ksh using 'joblog' to capture log ================= into $RUNDATA/joblog/jar100.log
vi joblog/jar100.log <-- review joblog (good for long jobs) ====================
uvlp12 joblog/jar100.log <-- can print to document problems ======================== for resolution by reponsible programmer
141210:205215:jar100: Begin Job=jar100 141210:205215:jar100: /home/mvstest1/testlibs/jcls/jar100.ksh 141210:205215:jar100: Arguments: jar100.ksh 141210:205215:jar100: RUNDATE=20141210 141210:205215:jar100: RUNLIBS=/home/mvstest1/testlibs 141210:205215:jar100: RUNDATA=/home/mvstest1/testdata 141210:205215:jar100: JTMP=jobtmp/jar100 141210:205215:jar100: SYOT=sysout/JAR100 141210:205215:jar100: LOGNAME=mvstest1 TESTPROD=T000 141210:205215:jar100: HOSTNAME=uvsoft4 141210:205215:jar100: ******** Begin Step S0010 car100 (#1) ******** 141210:205215:jar100: file: CUSTMAS=data1/ar.customer.master fsize=8K 141210:205215:jar100: file: NALIST=data1/ar.customer.nameadrs.list100 fsize=K 141210:205215:jar100: file: SYSOUT=sysout/JAR100/jar100_S0010_SYSOUT_141210_205215 fsize=K 141210:205215:jar100: Executing--> cobrun -F /home/mvstest1/testlibs/cblx/car100 141210:205215:JAR100: StepTimes: S0010 car100 Begun=20:52:15End=20:52:15 Elapsed=00:00:00 141210:205215:JAR100: Job Times: Begun=20:52:15 NormalEnd=20:52:15 Elapsed=00:00:00 141210:205215:jar100: JobEnd=Normal, StepsExecuted=1, LastStep=S0010
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
# joblog2 - run JCL/script with delay until any prior same script completes
# - creates logfile in subdir joblog/... with JCL/script name
# - appends _date_time stamp & return code
# (which changes name & allows waiting same jobnames to run)
# - writes the log file to joblog/jobname_yymmdd_hhmmss_RC
#
# Versions of joblog:
# joblog - names log file same as JCL/script with .log replacing .ksh
# - intended for test/debug
# joblog1 - appends date/time stamp to log file
# joblog2 - delays execution if same jobname already running
# - date/time stamps at end of job (allows waiting job to run)
#
jclksh="$1" # capture script filename with extension (jclname.ksh)
jcl=${jclksh%\.*} # drop extension .ksh from JCL/script filename
if [[ -f $RUNLIBS/jcls/$jclksh || -f $RUNLIBS/ksh/$jclksh ]]; then :
else echo "usage: joblog2 jclname.ksh [args]"
echo " =========================="
echo " - arg1 must be a script in the PATH"
exit; fi
#
# setup joblog directory pathname & make joblog subdir if not existing
#Jan08/2014 - ensure write to $RUNDATA (testdata) no matter where run
jld=$RUNDATA/joblog # setup joblog directory pathname
if [[ ! -d $jld ]]; then mkdir $jld; fi
#
# delay until any other same jobname script completes
stime=2; stotal=0
until [[ ! -f $jld/$jcl ]]
do echo "waiting $stime seconds for same-name job to complete"
sleep $stime
((stotal+=stime))
if [[ $stotal -gt 4 ]]; then stime=10; fi
if [[ $stotal -gt 60 ]]; then stime=60; fi
done
#
# create logfilename by appending date/time stamp
jcllog=${jcl}_$(date +%y%m%d_%H%M%S)
#
$jclksh $* 2>&1 | tee $jld/$jcl
#==============================
# "$*" will include all arguments spcfd after jclname.ksh
JCC=$? # capture return code to append on logfile name
# mv $jld/$jcl $jld/${jcllog}_$JCC
# - does not work because we get status of tee (vs $jclksh)
#
# change name of logfile by appending _date_time on JCL/script name
# - which will allow any waiting same-name job to run
mv $jld/$jcl $jld/$jcllog
#========================
exit 0
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
We will modify demo job jar100.ksh, inserting a pause so we can start a 2nd copy before 1st copy completes.
#1. Login --> /home/userxx
#2. cdl --> /home/userxx/testlibs
#3. vi jcls/jar100.ksh <-- edit jar100.ksh to insert following at line 34
==================
--> echo "enter to continue"; read reply #<-- wait for 2nd job to be run <--**
==================================== - see line 34 in script below
#!/bin/ksh
##JAR100 JOB 'TEST MVS2UNIX CONVERSION',DATE=20080323
export jobid2=jar100 JOBID2=JAR100; scriptpath="$0"; args="$*"
if [[ -z "$jobid1" ]]; then export jobid1=$jobid2; fi
for arg in $args; do if [[ "$arg" == *=* ]]; then export $arg; fi; done
integer JCC=0 SCC=0 LCC=0 # init step status return codes
autoload jobset51 jobend51 jobabend51 logmsg1 logmsg2 stepctl51
autoload exportfile exportgen0 exportgen1 exportgen2 exportgenall exportgenx
. $APPSADM/env/stub.ini #<-- for control-M (see notes in env/stub.ini)
jobset51 # call function for JCL/script initialization
goto
S0000=A
# * CUSTOMER NAME & ADDRESS LIST (WITH DATE & RPT HDNGS)
#1======================= begin step#S0010 CAR100 ========================
S0010=A
stepctl51; #calc StepTimes & check jobstop/goto end
goto
export JSTEP=S0010; ((XSTEP+=1)); SCC=0; LCC=0; alias goto="";
logmsg2 "******** Begin Step $JSTEP car100 (#$XSTEP) ********"
##STEP010 EXEC PGM=CAR100,PARM=2008
export PROGID=car100
export PARM="2008"
exportfile CUSTMAS data1/ar.customer.master
exportfile NALIST data1/ar.customer.nameadrs.list100
#3----------------------------------------------------------------------
cobrun $ANIM $RLX/car100
#4----------------------------------------------------------------------
LCC=$?; S0010C=$LCC; ((SCC|=LCC)); ((JCC|=LCC)); S0010R=1; alias goto="";
if ((S0010C != 0))
then logmsg2 "ERR: step#$JSTEP car100 abterm $SCC"
alias goto="<<S9900=\A"; fi
goto
#8======================================================================
echo "enter to continue"; read reply #<-- wait for 2nd job to be run <--**
#8======================================================================
#Note - End of job Normal & Abnormal code omitted to save space
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
This assumes you have performed the 'demo preparation' above (inserting a prompt to stop job 1, allow running job 2, then reply to job 1)
#1. Login --> /home/userxx
#2. joblog2 jar100.ksh <-- run 1st copy of jar100.ksh
==================
--> prompts "enter to continue" <-- waits for operator reply
#3. Switch to another screen
========================
- via ctl/alt/function 1-8 if on unix console
- or different putty session if access via Windows putty terminal emulator
- or could Login as a different user with same $RUNLIBS & $RUNDATA
#4. joblog2 jar100.ksh <-- run 2nd copy of jar100.ksh
================== - note following "waiting" displays
waiting 2 seconds for same-name job to complete
--- repeat wait 2 seconds 3 times ---
waiting 10 seconds for same-name job to complete
--- repeat wait 10 seconds 6 times ---
waiting 60 seconds for same-name job to complete
--- repeat wait 60 seconds until reply to 1st job ---
#5. Switch back to 1st screen
=========================
#6. Reply to 1st "enter to continue" prompt
=======================================
- 1st job completes & 2nd job starts
#7. Switch to 2nd screen
====================
#8. Reply to 2nd "enter to continue" prompt
=======================================
- 2nd job completes
#9. l joblog <-- list log files in joblog/... (date_time stamped)
========
#10. vi joblog/jar100* <-- inspect logfiles
=================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
The JCL converter converts mainframe GDG files to Vancouver Utility function equivalents exportgen0 & exportgen1. 'exportgen0' determines the most recent generation for reading & 'exportgen1' determines the next generation for writing. For example, here are the file conversions from step 2 of jgl230.jcl page '2E1'.
//GLTRANS DD DSN=GL.ACCOUNT.TRANS(+1),DISP=OLD
//GLMSOLD DD DSN=GL.ACCOUNT.MASTER(0),DISP=OLD
//GLMSNEW DD DSN=GL.ACCOUNT.MASTER(+1),DISP=(,CATLG,DELETE),...
exportgen1 +1 GLTRANS data1/gl.account.trans_ exportgen0 0 GLMSOLD data1/gl.account.master_ exportgen1 +1 GLMSNEW data1/gl.account.master_
Vancouver Utility GDG files are identified in JCL/scripts by a '_' trailing underscore. The actual datafiles (in $RUNDATA/data1/...) have 6 digits following the '_'. The number of generations maintained is controlled by gdg=... in a control file $GDGCTL/gdgctl51I.dat/.idx (Indexed file loaded from $GDGCTL/gdgctl51 text file)
JCL/scripts call function 'exportgen1' to determine next generation# to create. exportgen1 deletes oldest gdg#s depending on gdg=___ in $GDGCTL/gdgctl51I.dat/.idx - for example if gdg=3, adding a 4th file will delete the 1st file
gl.account.master_000001 gl.account.master_000002 <-- sample GDG file with 3 generations gl.account.master_000003
gl.account.master_999999 <-- max gen#, need to reset before reached - would take 3000 years to reach max gen# if 1 new generation per day - would take 3 years if you create 1000 new generations/day - generation#s will be automatically reset at 900000 by jobend51/gdgreset1
New generations are created in jobtmp/GDG/... & moved back to data1/... at successful End-Of-Job by function 'jobend51' 'jobend51' also tests for GDG#s > 900000 & calls script 'gdgreset1' to re-sequence gdg#s > 900000 back to 100001, 100002, etc
gl.account.master_900001 <-- gdg file with 3 gens > 900000 gl.account.master_900002 gl.account.master_900003
gl.account.master_100001 <-- gdg#s re-sequenced from 100001 gl.account.master_100002 gl.account.master_100003
'gdgreset1' called by 'jobend51' at EOJ will reset gdg#s for all members of 1 file whenever a new member is created with gdg# > 900000.
Also see 'gdgreset2' on page '7D1' to reset gdg#s for 'ALL' files in directory with GDG# > a specified threshold. You would run gdgreset2 in an offline batch mode when no jobs are running (to reset ALL files with gdg#s >threshold vs gdgreset1 which resets only 1 file when a new member created > 900000).
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
If you wish to verify that the GDG# reset works as advetised, you could do so as shown below. We assume that you have already run the JCL conversions in JCLcnv1demo.htm#Part_3 & executions in Part 4, particularly jgl230.ksh executed on page JCLcnv1demo.htm#4F5 (see console log displayed).
Here are 'only' the datafiles relevant to our test (step 2 of jgl230.ksh).
/home/uvadm/mvstest/testdata <-- test/demo demo data files supplied /home/userxx <-- can copy to your homedir :-----testdata <-- data dirs (defined as $RUNDATA in JCL/scripts) : :------ctl <-- control files for JCL convert & GDG control : : :-----gdgctl51 <-- GDG control file text version : : :-----gdgctl51I.dat <-- Indexed file to supply gdg=... limit : : :-----gdgctl51I.idx : :------data1 - datafiles supplied for test/demos : : :-----gl.account.master_900001 : : :-----gl.account.master_900002 <-- gen# changed to 900001,2,3 : : :-----gl.account.master_900003 : : :-----gl.account.tran1 : : :-----gl.account.trans_000001 : : :-----gl.account.trans_000002 : : :-----gl.account.trans_000003 : :------joblog <-- log files : :------jobtmp/GDG/... <-- new generations created here : : - moved back to data1/ at successful EOJ
#1. Login userxx --> $HOME (/home/userxx) #2. cdd --> $HOME/testdata
#3. change generation#s on gl.account.master_ to 900001,900002,900003 #3a. mv data1/gl.account.master_000001 data1/gl.account.master_900001 #3b. mv data1/gl.account.master_000002 data1/gl.account.master_900002 #3c. mv data1/gl.account.master_000003 data1/gl.account.master_900003
#4. testdatainit <-- clear joblog & jobtmp subdirs of prior runs
============ - also prompts to reload the GDG control file
#5. joblog jgl230.ksh <-- execute jgl230.ksh, with joblog to save
================= console messages in joblog/jgl230.log
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
150422:103638:jgl230: Begin Job=jgl230 150422:103638:jgl230: ... Step S0010 uvsort ... (omitted) 150422:103638:jgl230: ******** Begin Step S0020 cgl200 (#2) ******** 150422:103638:jgl230: gen+1 GLTRANS=/home/mvstest/testdata/data1/gl.account.trans_ gens=7 150422:103638:jgl230: gen0 GLMSOLD=data1/gl.account.master_900003 insize=16K 150422:103638:jgl230: gen+1 GLMSNEW=/home/mvstest/testdata/data1/gl.account.master_ gens=7 150422:103638:jgl230: Executing--> cobrun -F /home/mvstest/testlibs1/cblx/cgl200 150422:103638:JGL230: StepTimes: S0020 cgl200 Begun=10:36:38 End=10:36:38 Elapsed=00:00:00 150422:103638:jgl230: ... Step S0030 cgl100 ... (omitted) 150422:103638:jgl230: moving jobtmp/jgl230/GDG/files back to intended directories /home/mvstest/testdata/data1/gl.account.master_900004 /home/mvstest/testdata/data1/gl.account.trans_000004 gdgreset1 - script to resequence gdg#s for 1 file if gdg# > threshold gdgreset1 data1/gl.account.master_900001 ---> gl.account.master_100001 gdgreset1 data1/gl.account.master_900002 ---> gl.account.master_100002 gdgreset1 data1/gl.account.master_900003 ---> gl.account.master_100003 gdgreset1 data1/gl.account.master_900004 ---> gl.account.master_100004 gdgreset1 - 4 GDG files resequenced 150422:103638:JGL230: StepTimes: S0030 cgl100 Begun=10:36:38 End=10:36:38 Elapsed=00:00:00 150422:103638:JGL230: Job Times: Begun=10:36:38 NormalEnd=10:36:38 Elapsed=00:00:00 150422:103638:jgl230: JobEnd=Normal, StepsExecuted=3, LastStep=S0030
#6. ls -l data1 <-- display results
=========== - showing only files relevant to step 2 of jgl230.ksh
: : :-----gl.account.master_100001 : : :-----gl.account.master_100002 <-- gen#s reset to 100001,2,3,4 : : :-----gl.account.master_100003 : : :-----gl.account.master_100004 <-- new gen created by updt as90004 : : ; re-sequenced to 100004 by jobend/gdgreset1 : : :-----gl.account.tran1 : : :-----gl.account.trans_000001 : : :-----gl.account.trans_000002 : : :-----gl.account.trans_000003 : : :-----gl.account.trans_000004
| Note |
|
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
'gdgrest1' (above) resets gdg#s for 'ONE' file automatically, whenever gdg# 900000 is exceeded. 'gdgreset1' is called by function 'jobend51' which was inserted by the JCL converter at the end of all JCL/scripts.
'gdgreset2' (documented here) will reset GDG#s for 'ALL' files whose GDG#s exceed a specified threshold (suggested default 500000). You would run this batch job when no JCL/scripts are being executed.
We will demonstrate gdgreset2 using the same files as used above for gdgreset1. This is easy to do, but of course will show only the 1 file being reset (vs many that may be reset in a real situation).
'gdgreset2' creates a 'script' to copy/rename files with gdg#s > specified threshold. The script will copy all files to a 2nd directory (renaming if gdg# was > threshold). You can inspect before renaming the input as data1.old & the output dir back to data1.
#3. change generation#s on gl.account.master_ to 500001,500002,500003
#3a. mv data1/gl.account.master_000001 data1/gl.account.master_500001 #3b. mv data1/gl.account.master_000002 data1/gl.account.master_500002 #3c. mv data1/gl.account.master_000003 data1/gl.account.master_500003
#4a. mkdir data1_reset <-- make dir for output #4b. mkdir sf <-- make dir to receive output script
#5. uvcopy gdgreset2,fild1=indir,fild2=outdir,filo1=script-to-copy/resequence-gdgs
,uop=m______(Threshold),n______(ResequenceFrom)
==============================================================================
- command format above
- see actual command on the next page
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page
#5a. uvcopy gdgreset2,fild1=data1,fild2=data1_reset,filo1=sf/data1_reset.ksh
,uop=m500000,n100001
=======================================================================
#5b. uvcopy gdgreset2 <-- same as above, files default as shown above
================ - prompts allow change, as shown below
uop=m500000n100001 - option defaults
m500000 - default gdg# threshold for reset
n100001 - default gdg# to resequece from
User OPtion (uop) defaults = q1m500000n100001
null to accept or enter/override --> <-- null to accept defaults
did you create empty outdir (or remove files) -------> <-- null OK
data1 = default fild01 - null accept or reenter ? --> <-- null OK
data1_reset = default fild02 - accept or reenter ? --> <-- null OK
sf/data1.reset.ksh = default filo01 - accept ? --> <-- null OK
#6. vi sf/data1_reset.ksh <-- inspect script before execution
=====================
# gdgreset2 - reset generation numbers for Vancouver Utility GDG files
# 2015/04/21_16:52:46 InDir: data1 OutDir: data1_reset
# if gdg#s > 500000 - resequence from 100001
cp data1/gl.account.master_500001 data1_reset/gl.account.master_100001
cp data1/gl.account.master_500002 data1_reset/gl.account.master_100002
cp data1/gl.account.master_500003 data1_reset/gl.account.master_100003
cp data1/gl.account.trans_000001 data1_reset/
cp data1/gl.account.trans_000002 data1_reset/
cp data1/gl.account.trans_000003 data1_reset/
#7. sf/data1_reset.ksh <-- execute script to copy/rename files
================== from data1/... --> data1_reset/...
#8. ls -l data1_reset <-- inspect output results
=================
-rw-r--r--. 1 mvstest1 apps 13952 Apr 21 16:47 gl.account.master_100001
-rw-r--r--. 1 mvstest1 apps 13952 Apr 21 16:47 gl.account.master_100002
-rw-r--r--. 1 mvstest1 apps 13952 Apr 21 16:47 gl.account.master_100003
-rw-r--r--. 1 mvstest1 apps 1600 Apr 21 16:47 gl.account.trans_000001
-rw-r--r--. 1 mvstest1 apps 1600 Apr 21 16:47 gl.account.trans_000002
-rw-r--r--. 1 mvstest1 apps 1600 Apr 21 16:47 gl.account.trans_000003
#9a. mv data1 data1_old <-- rename current as old
===================
#9b. mv data1_reset data1 <-- rename newdir as current data1 dir
====================
Goto: Begin this doc , End this doc , Index this doc , Contents this library , UVSI Home-Page