Mainframes Tips And Tricks 6a5i6p
This document was ed by and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this report form. Report r6l17
Overview 4q3b3c
& View Mainframes Tips And Tricks as PDF for free.
More details 26j3b
- Words: 3,479
- Pages: 12
Mainframe Tips and Tricks
Mainframes Tips and Tricks Author: Ameya Manohar Last Modified: 26/05/06
Page 1 of 12
Mainframe Tips and Tricks
Preface This document is a collection of some tips that have been published in various manuals, books, websites and online forums. The primary motto of the document is to provide to mainframe application developers, tips which can save valuable time speed up development and enhancement activities.
Thanks and Warm Regards Ameya Manohar 25th May 2006
Page 2 of 12
Mainframe Tips and Tricks
1) While in EDIT mode, how to identify the changes that have been done since the last SAVE command was issued? Give “COMPARE SESSION” on the command line. 2) Normally a PDS or PS bears the ID of the creator/recent modifier. But is it possible to change these ID values without leaving a trace of ones own ID? It is very much possible… Entering ‘G’ in the command area against the PDS/PS throws a dialog box which allows you to change the ID values 3) How to identify the current TSO area from any ISPF screen Issue ‘SAREA’ on the command line. 4) Any command entered on the command line disappears after it is successfully executed. If someone has to issue the same command again then it either has to be retyped or one has to use specific PF keys. However there is one method which makes the command stay on the screen even after its successful execution. What is the method? Precede the commands with an ‘&’ Eg. &chg all ‘AAA’ ‘XXX’ After the execution, the command stays on the command line and does not disappear. This way you can enter the same command with minimum/without changes multiple times. 5) How to identify who is using a member currently, after you get “Member in Use” message, while trying to open a member in a PDS? Press F1 twice
6) Normally a search for in PDS is done by the first few characters of member name. But what if I want to search for with middle alphabets. Eg. Suppose there are some like MEDSJCL1, HYPEJCL2, TYPEGDG1, MEDSGDG2, TYPECBK1, MEDSCBK2, HYPECOB1 and BPMNCOB2. Now, how to find with GDG in their name? Use ISPF 4.12 option. Specify the dataset to be searched, pattern of the string, and specify the LANG type as COB. Select the BROWSE option. 7) How to find the last 10 datasets that were accessed by me? Use ISPF 3.4 From the menubar at the top, select REFLIST and select Option 1. 8) At times, the COMPARE macro can be used instead of option 3.13 as follows. It is usually helpful when the changes between the two datasets being compared are minimal.
Page 3 of 12
Mainframe Tips and Tricks
Open a DSN01 in Edit/View mode. On the command line issue COMP ‘DSN02’. If Only changes are to be listed issue COMP ‘DSN02’ EXC. 9) Ever heard of Auto Complete feature in Mainframes? Try this Use option 3.4 and enter KEYS on the command line. When the key list window pops up, set any free PF key to AUTOTYPE. Now save the key list and exit. Now type any dataset partially and press the assigned PF key. However, this method cannot retrieve names of partially typed VSAM or GDG files. 10) How to find a member without knowing the PDS to which it belongs and when only the high level qualifier is known. Using 3.4 list all PDS with the high level qualifier (Eg. ID.APPL.*) and issue MEMBER membername on the command line. 11) How to find out the last 25 COMMAND LINE commands issued by me? On the command line type RETP and hit enter. 12) I want to edit a member which has AAM as middle string in the PDS ID.APPL.MEM In the ISPF 3.4 listing of the PDS, on the command line issue S *AAM* E. This way one need not see the list of in the PDS. 13) Normally, to find the last edited member in a PDS, we issue a SORT CHA on the command line. However all these key strokes can be saved… Just place the cursor on the CHANGED header in 3.4 listing and hit enter. The will be sorted on that column. 14) Want to avoid 3.4 option? No problem!!! Issue EPDF ‘ID.APPL.PDS’ on the command line. The PS/PDS will open in EDIT mode. For View/Browse mode just append VIEW/BROWSE to the command.
15) Suppose while making changes to a program, I have to go out somewhere for a short while, and I leave without saving the changes I made. Now upon returning, how can I find out the changes made by me without getting out of the screen? Simply issue COMP * X on the command line.
16) Mainframe also allows one to view the contents of the clipboard. Don’t believe it? After a valid CUT command, use the command CUT DISPLAY. The clipboard manager dialog pops up with options to Edit/Browse the contents of the clipboard.
Page 4 of 12
Mainframe Tips and Tricks
17) Suppose there is a production file with 10,000 records and you want to create a test file with the first 10 and the last 10 records from the production file. How to do this without having to cut/paste twice? Quite simple! Cut the first 10 line and issue CUT APPEND command. Then cut the last 10 lines and issue CUT APPEND again. Upon PASTE you will get all 20 lines. 18) Tired of typing the often used datasets? Go to 3.4 option and enter the KEYS command on the command line. In the key list, assign any free PF key to NRETRIEVE and save and exit. Now, whenever you are on 3.4 option, pressing the assigned PF key will bring up the last 30 accessed datasets one by one.
19) What if I want to search a string ‘ABCXYZ’ in the of a PDS? There may be other strings ABCXYZ but I only want to locate those which have quotes. Use “’ABCXYZ’” in option 3.14 20) I want to compare two files, but I want to compare them based on fields in a particular column. For Eg. I want to compare file 1 column 1 to 10 with file 2 column 30 to 40. How do I do it? Go to option 3.13. On the same screen, on the command line, issue E. In the space, type CMPCOLN 30:40 on one line and CMPCOLO 1:10 on next line. Exit using PF3. In the NEW dataset, type file2 and in the OLD dataset, type file1. 21) How to search for a member in multiple PDS? Using 3.4 list all the possible datasets. On the command line type M MEMBERNAME. This does not work with migrated datasets. 22) Is there a command similar to WORDWRAP on mainframes? Of course! Open the dataset in edit mode. On the command line type TFXX (where XX is the no. of characters that are to be wrapped) and hit enter. All the subsequent lines will wrap to XX characters per line. 23) From the ISPF 3.4 listing, it is possible to directly EDIT/BROWSE/VIEW a member without listing all the and then selecting the particular member. On the command area besides the PDS listing, issue E/MEMBERNAME 24) If, due to some reason, you want to delete a protected (using RETPD option) dataset, use the IDCAMS DELETE command with PURGE option as follows:
Page 5 of 12
Mainframe Tips and Tricks
//SYSIN DD * DELETE ID.APPL.PDS – PURGE /* 25) How to execute any program from any library using referback? //STEP01 EXEC PGM=IEFBR14 //DDNAME01 DD DSN=PDS.APPL.LOADLIB(COBPGM),DISP=SHR //STEP02 EXEC PGM= *.STEP01.DDNAME01
26) How does the system differentiate between JES3 statements and JCL comments (both start with //*) If the 4th column contains a JES3 statement, the line is considered to be a JES3 line, otherwise it is treated as a normal JCL statement. 27) We have a JES2 and JES3. But was there ever any JES1? OS/VS1 ed only one job entry subsystem, JES1.
28) There is a JCL with 10 steps but I want to execute only the 5th step. How can I do this? One way would be to use RESTART=STEP05. But this will also execute the subsequent steps. So on the job card, code the COND parameter as follows: RESTART=STEP05,COND(0,LE) 29) Can catalog procedure have a PEND statement? Yes. PEND can be used even with catalog procedures. 30) How much data can be stored on a TAPE? A logical tape reel can hold 170 MB where as 10GB of data can be accommodated on a tape cartridge. 31) What is the origin of the NULL statement in JCL? During the old days of punched cards, a null coded on distinctive colored cards was used to help operators separate card decks. Thus it helped to identify end of a job card. 32) Why do JCL statement have a ‘//’ in column 1 and 2? This also dates back to the punched cards era. If the cards were accidentally put back to front, hen the slashes won’t be read first so the job (card deck) would be rejected. Thus ‘//’ was used to ensure that the card deck is placed in the reader correctly.
Page 6 of 12
Mainframe Tips and Tricks
33) How to specify a dummy VSAM dataset in a JCL? Use AMP=’AMORG’ parameter on the dummy DD statement. 34) What is the difference between TIME=MAXIMUM and TIME=NOLIMT? TIME=MAXIMUM allows the job to run for 357912 minutes while TIME=NOLIMIT allows the job to run for unlimited amount of time (The job can remain in wait status for more than the installation time limit). 35) I want to delete a member in a PDS but someone is using it. How do I delete it? //STEP01 EXEC PGM=IDCAMS //DDNAME01 DD DSN=PDS,DISP=SHR //SYSIN DD * ALLOC DDN(INFILE) SHR REUSE DSN(‘PDS’) DELETE PDS(MEMBER) FILE(DDNAME01) /* 36) &SYSUID is the only system symbol that can be used directly in a batch job. 37) I want to override 2 DD statements in my PROC from my JCL. One statement is in step 10 and the other is in step 40. What if I specify the override as follows: //STEP040.DDNM01 DD DSN=DATASET01 //STEP010.DDNM02 DD DSN=DATASET02 z/OS JCL will still process it. It will override irrespective of the specifying order. 38) What care needs to be taken while accessing an ESDS file in a COBOL program? For ESDS files, the syntax for the SELECT clause is modified a little by prefixing an AS- to DDNAMES. Eg. SELECT FILE01 ASSIGN TO AS-DDNAME otherwise an S213 abend occurs while opening the dataset. 39) When writing a COBOL program that is going to be used as a CICS application program, the DYNAM compiler option should not be used. 40) MERGE statement in COBOL can have a USING OUTPUT PROCEDURE but not a USING INPUT PROCEDURE.
41) How does a COBOL compiler differentiate between VSAM datasets and Flat files? When the ORGANIZATION clause is specified, the COBOL compiler interprets it as a VSAM dataset. Hence the clause should not be used for non-VSAM files. 42) Using an odd no of digits for packed decimal (COMP3) is 5% to 20% faster than using an even no of digits.
Page 7 of 12
Mainframe Tips and Tricks
43) As per the COBOL Performance tuning manual, using a COMP subscript to address a table is 30% slower than using indexes, using COMP3 is 300% slower, while a DISPLAY subscript is slower than indexes by a whopping 450% 44) As per the rule of thumb, for a table with less than 50 entries, use SEARCH option. For a table with greater than 50 entries use SEARCH ALL. 45) In COBOL, why do we READ FILE but WRITE RECORD? We READ FILE because we don’t know if there actually is a record to be read and (for variable and undefined length files) how long he next record will be. However, while writing to a file, we know both the above in advance. 46) Using OPEN OUTPUT to load a VSAM file significantly improves the performance of the program. Using OPEN I-O and OPEN EXTEND has a negative impact on the programs performance. 47) Repetitive use of INITIALIZE statement should be avoided. 48) An inline PERFORM can be used instead of a SEARCH when the table contains less than 20 elements. 49) To generate a complete listing of the compiler diagnostic messages, with their explanations, compile a program with the program-id ERRMSG 50) When processing KSDS or RRDS VSAM files, when DELETE statement is used, the file must be opened in I-O mode. 51) Taking a constant expression out of the loop speeds up the program with absolutely no ill effects. 52) When INITIALIZE verb is used to initialize group variables, group elements which are FILLERS are not initialized. 53) How can we test the return code of an internal sort? The return code of an internal sort is stored in a special SORT-RETURN. 54) In general, it is advantageous to use COMP for numeric and COMP3 for decimal data. 55) A better way of initializing a record or group item is INITIALIZE PV-RECORD REPLACING ALPHANUMERIC DATA BY SPACES NUMERIC DATA BY ZEROES 56) SEARCH ALL can only be used to test an equal condition.
Page 8 of 12
Mainframe Tips and Tricks
57) One way to make the COBOL program more maintenance friendly is to code the ASSIGN clause as follows: SELECT FILE01 ASSIGN TO XX-Z-DDNAME where XX – Device Class (Tape, DASD) Z – Method of Organization (Sequential, Indexed, Random) 58) When using INPUT/OUTPUT procedure with SORT, we RELEASE record and RETURN file.
59) When dealing with VSAM files, one can speed up the applications by increasing the VSAM buffers. In the DD statement, code AMP parameters = BUFNI(index), BUFND(data), BUFSP(virtual storage). For sequential access, increase the number of data buffers. For random access, increase index buffers. 60) IDCAMS utility cannot be executed without two DD statements. They are SYSPRINT and SYSIN. 61) Why does FREESPACE(100,100) cause one record to be written for each CI, and one CI for each CA? This is because VSAM always loads the first record before taking FREESPACE option into consideration. 62) DEFINE PATH creates an alias for the VSAM base cluster while DEFINE ALIAS creates an alias for non VSAM datasets 63) CICS, DB2 and IMS can process records in existing VSAM datasets, but cannot create the dataset. Even in COBOL, VSAM datasets can be processed only after they are created using access method services. 64) How to specify a dummy VSAM dataset in a JCL? Use AMP=’AMORG’ parameter 65) What are the 5 IDCAMS commands that must be executed before the records in a KSDS can be processed through an alternate key? DEFINE CLUSTER, REPRO, DEFINE ALTERNATEINDEX, DEFINE PATH, BLDINDEX 66) How to make a VSAM dataset read only? By using INHIBIT option along with ALTER. 67) The VSAM equivalent of a temporary dataset is to create a VSAM dataset with REUSE option.
Page 9 of 12
Mainframe Tips and Tricks
68) How should a VSAM file be defined so that it can be accessed both in CICS and BATCH environment? By using SHAREOPTION(2,3). 69) The easiest way to find bad data in a file is to use the Find command in FILEAID. Open the file in FILEAID, XREF with the copybook, and use FMT mode. Now issue F /FLD INVALID where FLD = Either filed name or field number. For Eg. In a file containing EMP NAME, EMP NO, SALARY as fields, to find the bad data in EMP NO, use F /2 INVALID. 70) To swap between BROWSE and EDIT modes in FILEAID without closing the file use the GO command. Eg GO EDIT to go from browse to edit mode. 71) To determine the compile date of a load module use file aid as follows: Use file aid 3.1 option. In the library utility type the library name and type ‘ a’ next to the member name. TDATE is the compile date. 72) A cool way to count no. of records in a VSAM file is to use the option 3.8 in FILEAID and type in TALLY at the prompt. To exit type END
73) To have “Freeze Panes” option of MS Excel on mainframes, open FILE-AID. Use VFMT format and issue HOLD command. Eg. HOLD 1-3,6. This freezes columns 1,2,3 and 6.
74) Unlike ISPF, FILE_AID allows the use of FIND command with relational operators like EQ,NE,GT,GE, etc… Eg. FIND GT ‘100’ 25 ALL will locate all the lines which have value greater than 100 in the 25th column. 75) How to display only certain fields of a very large file? Open the dataset with FILE-AID in VFMT mode, and issue the command DISPLAY X Y Z only where X,Y,Z are column nos to be displayed.
76) I have a file with about 10 million records out of which some records have bad data in one particular field. How do I eliminate the records in a fast efficient manner? Execute FILEAID in batch mode with $$DDNAME DROP IF=(field-position,fieldlength,NEP) where NEP = Not equal to pack considering the field in error is a COMP3 variable. 77) Using FILE-AID COMPARE, one can set percentage amount for numeric fields that determines “close enough” matches. When the “Formatted Criteria” table appears after the 2 files to be compared are specified, use the “T” command. This brings up a Tolerance Value/Percentage Specification dialog where one can set a positive or negative value for a numeric field.
Page 10 of 12
Mainframe Tips and Tricks
78) If you SORT/MERGE with COBOL (internal sort), the compiler automatically generates a SORT/MERGE control statements with the correct DFSORT description for the specified COBOL fields 79) To replace all low values in a file with spaces in a FB dataset with LRECL of 80, use ALTSEQ CODE=(0040) OUTREC FIELDS=(1,80,TRAN=ALTSEQ) 80) One of my project requirements is that in one step, the contents of a VSAM file have to be copied to a flat file and the flat file is to be processed further. However, sometimes the VSAM dataset may be empty in which case, my job abends. How to prevent this? Use PARM=’VSAMEMT=YES’ on the EXEC statement of the job. 81) One column in SDSF that we can and should bring to the front position is the MAX-RC column. It shows the completion code or the abend code of the job. This can be done by issuing the command “arr ?” and using / and // to move the columns. 82) I have some 100 odd jobs in my spool. How to view only those jobs which have ABENDED? To obtain such a listing use the FILTER command as follows: FIL MAX AB* FIL MAX EQ ‘RC 0000’ will show successfully completed jobs. 83) When a job is selected in the spool using the S option, the entire job is selected without the break-up output groups like JESMSGLG, JESJCL, SYSPRINT, etc… But using the commands ‘N’ and ‘P’ allows one to navigate between various output groups. 84) Using CALL instead of EXEC CICS LINK improves the total U time by about 15 to 20%. This is because EXEC CICS LINK establishes new run unit, sets up the environment, and searches the CICS table thereby executing about 1400 machine code instructions. On the other hand, CALL executes about 12 machine code instructions. 85) CICS was formerly called PUCICS (Public Utilities Customer Information Control System). 86) What is the difference between EIBTIME and ASKTIME? ASKTIME requests for the current date and time while EIBTIME populates the date and time of the task initiation. 87) EXEC CICS ASSIGN can be used to access system values such as date, time, id etc. and store in a variable.
Page 11 of 12
Mainframe Tips and Tricks
88) How to determine the number of records deleted, when doing a mass delete using a generic key? By using the NUMREC(data-area) option with the generic key. Upon completion of the delete the data area will contain the number of records that were deleted. 89) Usage of ‘SET CURRENT TIMESTAMP’ (DB2) is more taxing on the resources as compared to ‘MOVE CURRENT-DATE TO…’ (MVS) 90) Avoid s that involve more than two tables. Break them up into multiple SQL statements. 91) When using CURSOR for read-only, use the FOR FETCH ONLY clause in SQL. 92) To get only first N rows using a query SELECT * FROM table-name WHERE condition FETCH FIRST/LAST N ROWS ONLY WITH UR; 93) A is always preferred over sub queries. 94) To select the duplicate rows from a table SELECT columns FROM table-name WHERE condition GROUP BY columns HAVING COUNT(*) > 1 95) For SQL declaration of DECIMAL(n,m) the COBOL equivalent is PIC S9(nm)v9(m) COMP-3.
End of Document
Page 12 of 12
Mainframes Tips and Tricks Author: Ameya Manohar Last Modified: 26/05/06
Page 1 of 12
Mainframe Tips and Tricks
Preface This document is a collection of some tips that have been published in various manuals, books, websites and online forums. The primary motto of the document is to provide to mainframe application developers, tips which can save valuable time speed up development and enhancement activities.
Thanks and Warm Regards Ameya Manohar 25th May 2006
Page 2 of 12
Mainframe Tips and Tricks
1) While in EDIT mode, how to identify the changes that have been done since the last SAVE command was issued? Give “COMPARE SESSION” on the command line. 2) Normally a PDS or PS bears the ID of the creator/recent modifier. But is it possible to change these ID values without leaving a trace of ones own ID? It is very much possible… Entering ‘G’ in the command area against the PDS/PS throws a dialog box which allows you to change the ID values 3) How to identify the current TSO area from any ISPF screen Issue ‘SAREA’ on the command line. 4) Any command entered on the command line disappears after it is successfully executed. If someone has to issue the same command again then it either has to be retyped or one has to use specific PF keys. However there is one method which makes the command stay on the screen even after its successful execution. What is the method? Precede the commands with an ‘&’ Eg. &chg all ‘AAA’ ‘XXX’ After the execution, the command stays on the command line and does not disappear. This way you can enter the same command with minimum/without changes multiple times. 5) How to identify who is using a member currently, after you get “Member in Use” message, while trying to open a member in a PDS? Press F1 twice
6) Normally a search for in PDS is done by the first few characters of member name. But what if I want to search for with middle alphabets. Eg. Suppose there are some like MEDSJCL1, HYPEJCL2, TYPEGDG1, MEDSGDG2, TYPECBK1, MEDSCBK2, HYPECOB1 and BPMNCOB2. Now, how to find with GDG in their name? Use ISPF 4.12 option. Specify the dataset to be searched, pattern of the string, and specify the LANG type as COB. Select the BROWSE option. 7) How to find the last 10 datasets that were accessed by me? Use ISPF 3.4 From the menubar at the top, select REFLIST and select Option 1. 8) At times, the COMPARE macro can be used instead of option 3.13 as follows. It is usually helpful when the changes between the two datasets being compared are minimal.
Page 3 of 12
Mainframe Tips and Tricks
Open a DSN01 in Edit/View mode. On the command line issue COMP ‘DSN02’. If Only changes are to be listed issue COMP ‘DSN02’ EXC. 9) Ever heard of Auto Complete feature in Mainframes? Try this Use option 3.4 and enter KEYS on the command line. When the key list window pops up, set any free PF key to AUTOTYPE. Now save the key list and exit. Now type any dataset partially and press the assigned PF key. However, this method cannot retrieve names of partially typed VSAM or GDG files. 10) How to find a member without knowing the PDS to which it belongs and when only the high level qualifier is known. Using 3.4 list all PDS with the high level qualifier (Eg. ID.APPL.*) and issue MEMBER membername on the command line. 11) How to find out the last 25 COMMAND LINE commands issued by me? On the command line type RETP and hit enter. 12) I want to edit a member which has AAM as middle string in the PDS ID.APPL.MEM In the ISPF 3.4 listing of the PDS, on the command line issue S *AAM* E. This way one need not see the list of in the PDS. 13) Normally, to find the last edited member in a PDS, we issue a SORT CHA on the command line. However all these key strokes can be saved… Just place the cursor on the CHANGED header in 3.4 listing and hit enter. The will be sorted on that column. 14) Want to avoid 3.4 option? No problem!!! Issue EPDF ‘ID.APPL.PDS’ on the command line. The PS/PDS will open in EDIT mode. For View/Browse mode just append VIEW/BROWSE to the command.
15) Suppose while making changes to a program, I have to go out somewhere for a short while, and I leave without saving the changes I made. Now upon returning, how can I find out the changes made by me without getting out of the screen? Simply issue COMP * X on the command line.
16) Mainframe also allows one to view the contents of the clipboard. Don’t believe it? After a valid CUT command, use the command CUT DISPLAY. The clipboard manager dialog pops up with options to Edit/Browse the contents of the clipboard.
Page 4 of 12
Mainframe Tips and Tricks
17) Suppose there is a production file with 10,000 records and you want to create a test file with the first 10 and the last 10 records from the production file. How to do this without having to cut/paste twice? Quite simple! Cut the first 10 line and issue CUT APPEND command. Then cut the last 10 lines and issue CUT APPEND again. Upon PASTE you will get all 20 lines. 18) Tired of typing the often used datasets? Go to 3.4 option and enter the KEYS command on the command line. In the key list, assign any free PF key to NRETRIEVE and save and exit. Now, whenever you are on 3.4 option, pressing the assigned PF key will bring up the last 30 accessed datasets one by one.
19) What if I want to search a string ‘ABCXYZ’ in the of a PDS? There may be other strings ABCXYZ but I only want to locate those which have quotes. Use “’ABCXYZ’” in option 3.14 20) I want to compare two files, but I want to compare them based on fields in a particular column. For Eg. I want to compare file 1 column 1 to 10 with file 2 column 30 to 40. How do I do it? Go to option 3.13. On the same screen, on the command line, issue E. In the space, type CMPCOLN 30:40 on one line and CMPCOLO 1:10 on next line. Exit using PF3. In the NEW dataset, type file2 and in the OLD dataset, type file1. 21) How to search for a member in multiple PDS? Using 3.4 list all the possible datasets. On the command line type M MEMBERNAME. This does not work with migrated datasets. 22) Is there a command similar to WORDWRAP on mainframes? Of course! Open the dataset in edit mode. On the command line type TFXX (where XX is the no. of characters that are to be wrapped) and hit enter. All the subsequent lines will wrap to XX characters per line. 23) From the ISPF 3.4 listing, it is possible to directly EDIT/BROWSE/VIEW a member without listing all the and then selecting the particular member. On the command area besides the PDS listing, issue E/MEMBERNAME 24) If, due to some reason, you want to delete a protected (using RETPD option) dataset, use the IDCAMS DELETE command with PURGE option as follows:
Page 5 of 12
Mainframe Tips and Tricks
//SYSIN DD * DELETE ID.APPL.PDS – PURGE /* 25) How to execute any program from any library using referback? //STEP01 EXEC PGM=IEFBR14 //DDNAME01 DD DSN=PDS.APPL.LOADLIB(COBPGM),DISP=SHR //STEP02 EXEC PGM= *.STEP01.DDNAME01
26) How does the system differentiate between JES3 statements and JCL comments (both start with //*) If the 4th column contains a JES3 statement, the line is considered to be a JES3 line, otherwise it is treated as a normal JCL statement. 27) We have a JES2 and JES3. But was there ever any JES1? OS/VS1 ed only one job entry subsystem, JES1.
28) There is a JCL with 10 steps but I want to execute only the 5th step. How can I do this? One way would be to use RESTART=STEP05. But this will also execute the subsequent steps. So on the job card, code the COND parameter as follows: RESTART=STEP05,COND(0,LE) 29) Can catalog procedure have a PEND statement? Yes. PEND can be used even with catalog procedures. 30) How much data can be stored on a TAPE? A logical tape reel can hold 170 MB where as 10GB of data can be accommodated on a tape cartridge. 31) What is the origin of the NULL statement in JCL? During the old days of punched cards, a null coded on distinctive colored cards was used to help operators separate card decks. Thus it helped to identify end of a job card. 32) Why do JCL statement have a ‘//’ in column 1 and 2? This also dates back to the punched cards era. If the cards were accidentally put back to front, hen the slashes won’t be read first so the job (card deck) would be rejected. Thus ‘//’ was used to ensure that the card deck is placed in the reader correctly.
Page 6 of 12
Mainframe Tips and Tricks
33) How to specify a dummy VSAM dataset in a JCL? Use AMP=’AMORG’ parameter on the dummy DD statement. 34) What is the difference between TIME=MAXIMUM and TIME=NOLIMT? TIME=MAXIMUM allows the job to run for 357912 minutes while TIME=NOLIMIT allows the job to run for unlimited amount of time (The job can remain in wait status for more than the installation time limit). 35) I want to delete a member in a PDS but someone is using it. How do I delete it? //STEP01 EXEC PGM=IDCAMS //DDNAME01 DD DSN=PDS,DISP=SHR //SYSIN DD * ALLOC DDN(INFILE) SHR REUSE DSN(‘PDS’) DELETE PDS(MEMBER) FILE(DDNAME01) /* 36) &SYSUID is the only system symbol that can be used directly in a batch job. 37) I want to override 2 DD statements in my PROC from my JCL. One statement is in step 10 and the other is in step 40. What if I specify the override as follows: //STEP040.DDNM01 DD DSN=DATASET01 //STEP010.DDNM02 DD DSN=DATASET02 z/OS JCL will still process it. It will override irrespective of the specifying order. 38) What care needs to be taken while accessing an ESDS file in a COBOL program? For ESDS files, the syntax for the SELECT clause is modified a little by prefixing an AS- to DDNAMES. Eg. SELECT FILE01 ASSIGN TO AS-DDNAME otherwise an S213 abend occurs while opening the dataset. 39) When writing a COBOL program that is going to be used as a CICS application program, the DYNAM compiler option should not be used. 40) MERGE statement in COBOL can have a USING OUTPUT PROCEDURE but not a USING INPUT PROCEDURE.
41) How does a COBOL compiler differentiate between VSAM datasets and Flat files? When the ORGANIZATION clause is specified, the COBOL compiler interprets it as a VSAM dataset. Hence the clause should not be used for non-VSAM files. 42) Using an odd no of digits for packed decimal (COMP3) is 5% to 20% faster than using an even no of digits.
Page 7 of 12
Mainframe Tips and Tricks
43) As per the COBOL Performance tuning manual, using a COMP subscript to address a table is 30% slower than using indexes, using COMP3 is 300% slower, while a DISPLAY subscript is slower than indexes by a whopping 450% 44) As per the rule of thumb, for a table with less than 50 entries, use SEARCH option. For a table with greater than 50 entries use SEARCH ALL. 45) In COBOL, why do we READ FILE but WRITE RECORD? We READ FILE because we don’t know if there actually is a record to be read and (for variable and undefined length files) how long he next record will be. However, while writing to a file, we know both the above in advance. 46) Using OPEN OUTPUT to load a VSAM file significantly improves the performance of the program. Using OPEN I-O and OPEN EXTEND has a negative impact on the programs performance. 47) Repetitive use of INITIALIZE statement should be avoided. 48) An inline PERFORM can be used instead of a SEARCH when the table contains less than 20 elements. 49) To generate a complete listing of the compiler diagnostic messages, with their explanations, compile a program with the program-id ERRMSG 50) When processing KSDS or RRDS VSAM files, when DELETE statement is used, the file must be opened in I-O mode. 51) Taking a constant expression out of the loop speeds up the program with absolutely no ill effects. 52) When INITIALIZE verb is used to initialize group variables, group elements which are FILLERS are not initialized. 53) How can we test the return code of an internal sort? The return code of an internal sort is stored in a special SORT-RETURN. 54) In general, it is advantageous to use COMP for numeric and COMP3 for decimal data. 55) A better way of initializing a record or group item is INITIALIZE PV-RECORD REPLACING ALPHANUMERIC DATA BY SPACES NUMERIC DATA BY ZEROES 56) SEARCH ALL can only be used to test an equal condition.
Page 8 of 12
Mainframe Tips and Tricks
57) One way to make the COBOL program more maintenance friendly is to code the ASSIGN clause as follows: SELECT FILE01 ASSIGN TO XX-Z-DDNAME where XX – Device Class (Tape, DASD) Z – Method of Organization (Sequential, Indexed, Random) 58) When using INPUT/OUTPUT procedure with SORT, we RELEASE record and RETURN file.
59) When dealing with VSAM files, one can speed up the applications by increasing the VSAM buffers. In the DD statement, code AMP parameters = BUFNI(index), BUFND(data), BUFSP(virtual storage). For sequential access, increase the number of data buffers. For random access, increase index buffers. 60) IDCAMS utility cannot be executed without two DD statements. They are SYSPRINT and SYSIN. 61) Why does FREESPACE(100,100) cause one record to be written for each CI, and one CI for each CA? This is because VSAM always loads the first record before taking FREESPACE option into consideration. 62) DEFINE PATH creates an alias for the VSAM base cluster while DEFINE ALIAS creates an alias for non VSAM datasets 63) CICS, DB2 and IMS can process records in existing VSAM datasets, but cannot create the dataset. Even in COBOL, VSAM datasets can be processed only after they are created using access method services. 64) How to specify a dummy VSAM dataset in a JCL? Use AMP=’AMORG’ parameter 65) What are the 5 IDCAMS commands that must be executed before the records in a KSDS can be processed through an alternate key? DEFINE CLUSTER, REPRO, DEFINE ALTERNATEINDEX, DEFINE PATH, BLDINDEX 66) How to make a VSAM dataset read only? By using INHIBIT option along with ALTER. 67) The VSAM equivalent of a temporary dataset is to create a VSAM dataset with REUSE option.
Page 9 of 12
Mainframe Tips and Tricks
68) How should a VSAM file be defined so that it can be accessed both in CICS and BATCH environment? By using SHAREOPTION(2,3). 69) The easiest way to find bad data in a file is to use the Find command in FILEAID. Open the file in FILEAID, XREF with the copybook, and use FMT mode. Now issue F /FLD INVALID where FLD = Either filed name or field number. For Eg. In a file containing EMP NAME, EMP NO, SALARY as fields, to find the bad data in EMP NO, use F /2 INVALID. 70) To swap between BROWSE and EDIT modes in FILEAID without closing the file use the GO command. Eg GO EDIT to go from browse to edit mode. 71) To determine the compile date of a load module use file aid as follows: Use file aid 3.1 option. In the library utility type the library name and type ‘ a’ next to the member name. TDATE is the compile date. 72) A cool way to count no. of records in a VSAM file is to use the option 3.8 in FILEAID and type in TALLY at the prompt. To exit type END
73) To have “Freeze Panes” option of MS Excel on mainframes, open FILE-AID. Use VFMT format and issue HOLD command. Eg. HOLD 1-3,6. This freezes columns 1,2,3 and 6.
74) Unlike ISPF, FILE_AID allows the use of FIND command with relational operators like EQ,NE,GT,GE, etc… Eg. FIND GT ‘100’ 25 ALL will locate all the lines which have value greater than 100 in the 25th column. 75) How to display only certain fields of a very large file? Open the dataset with FILE-AID in VFMT mode, and issue the command DISPLAY X Y Z only where X,Y,Z are column nos to be displayed.
76) I have a file with about 10 million records out of which some records have bad data in one particular field. How do I eliminate the records in a fast efficient manner? Execute FILEAID in batch mode with $$DDNAME DROP IF=(field-position,fieldlength,NEP) where NEP = Not equal to pack considering the field in error is a COMP3 variable. 77) Using FILE-AID COMPARE, one can set percentage amount for numeric fields that determines “close enough” matches. When the “Formatted Criteria” table appears after the 2 files to be compared are specified, use the “T” command. This brings up a Tolerance Value/Percentage Specification dialog where one can set a positive or negative value for a numeric field.
Page 10 of 12
Mainframe Tips and Tricks
78) If you SORT/MERGE with COBOL (internal sort), the compiler automatically generates a SORT/MERGE control statements with the correct DFSORT description for the specified COBOL fields 79) To replace all low values in a file with spaces in a FB dataset with LRECL of 80, use ALTSEQ CODE=(0040) OUTREC FIELDS=(1,80,TRAN=ALTSEQ) 80) One of my project requirements is that in one step, the contents of a VSAM file have to be copied to a flat file and the flat file is to be processed further. However, sometimes the VSAM dataset may be empty in which case, my job abends. How to prevent this? Use PARM=’VSAMEMT=YES’ on the EXEC statement of the job. 81) One column in SDSF that we can and should bring to the front position is the MAX-RC column. It shows the completion code or the abend code of the job. This can be done by issuing the command “arr ?” and using / and // to move the columns. 82) I have some 100 odd jobs in my spool. How to view only those jobs which have ABENDED? To obtain such a listing use the FILTER command as follows: FIL MAX AB* FIL MAX EQ ‘RC 0000’ will show successfully completed jobs. 83) When a job is selected in the spool using the S option, the entire job is selected without the break-up output groups like JESMSGLG, JESJCL, SYSPRINT, etc… But using the commands ‘N’ and ‘P’ allows one to navigate between various output groups. 84) Using CALL instead of EXEC CICS LINK improves the total U time by about 15 to 20%. This is because EXEC CICS LINK establishes new run unit, sets up the environment, and searches the CICS table thereby executing about 1400 machine code instructions. On the other hand, CALL executes about 12 machine code instructions. 85) CICS was formerly called PUCICS (Public Utilities Customer Information Control System). 86) What is the difference between EIBTIME and ASKTIME? ASKTIME requests for the current date and time while EIBTIME populates the date and time of the task initiation. 87) EXEC CICS ASSIGN can be used to access system values such as date, time, id etc. and store in a variable.
Page 11 of 12
Mainframe Tips and Tricks
88) How to determine the number of records deleted, when doing a mass delete using a generic key? By using the NUMREC(data-area) option with the generic key. Upon completion of the delete the data area will contain the number of records that were deleted. 89) Usage of ‘SET CURRENT TIMESTAMP’ (DB2) is more taxing on the resources as compared to ‘MOVE CURRENT-DATE TO…’ (MVS) 90) Avoid s that involve more than two tables. Break them up into multiple SQL statements. 91) When using CURSOR for read-only, use the FOR FETCH ONLY clause in SQL. 92) To get only first N rows using a query SELECT * FROM table-name WHERE condition FETCH FIRST/LAST N ROWS ONLY WITH UR; 93) A is always preferred over sub queries. 94) To select the duplicate rows from a table SELECT columns FROM table-name WHERE condition GROUP BY columns HAVING COUNT(*) > 1 95) For SQL declaration of DECIMAL(n,m) the COBOL equivalent is PIC S9(nm)v9(m) COMP-3.
End of Document
Page 12 of 12
Related Documents 171j1w
Mainframes Tips And Tricks 6a5i6p
November 2019 83
Netezza Tips And Tricks 1s3k43
April 2022 0
Tips And Tricks 535326
February 2021 0
Travian Tips And Tricks 2r4628
October 2021 0
Microstation Tips And Tricks 6j6g1p
July 2021 0