Setup for IT Service Vision Server

Reporting on DCOLLECT Data

Follow this path from the main menu:

   Reports -> Supplied Reports

IT Service Vision displays the supplied report definitions that are based on the tables in your PDB. To see the supplied report definitions for DCOLLECT data, follow this path from the list of supplied report definitions:

   Edit -> Sort -> Table

IT Service Vision displays the list sorted by table name. Scroll to the report definitions whose table names begin with the characters 'XDC'. (The table names begin 'XDC' and the corresponding supplied report definitions begin 'DC'.) See Shared Appendix 7: Supplied Report Naming Convention and MVS Appendix 1: Tables and Variables Naming Convention for more information.

Report definitions that are based on groups and provide summaries of disk usage or of dataset space require additional setup before they can be used to generate reports. (These report definitions are listed later in this section.) The setup must be done at your site because the setup involves grouping volumes and datasets appropriately for your site.

You must provide formula variables to define these groups based on the disk and dataset naming conventions at your site. Formula variables are variables whose values are calculated (from the values of other variables and from constants) when you access the data. Formula variables do not occupy disk space in the PDB. The following examples show you various automated and "manual" methods that you can use to define the formula variables.

You may want to include the code that defines the formula variables (including the creation of a format, when applicable) at the beginning of the daily job production job so that the values of the formula variables are always current. Check the "manual" definitions of formula variables periodically and update them as necessary to keep them current.

Note: Some of these techniques make use of the search path for formats. In IT Service Vision on the server host, the default search path for formats is

   DICTLIB.CPFMTS ->
   SITELIB.CPFMTS ->
   PGMLIB.CPFMTS ->
   WORK.FORMATS ->
   LIBRARY.FORMATS ->

Reporting (via XUSEGRP) on Disk Usage

The supplied reports named DCSPPIES, DCSPUSES, DCSPCAUS, and DCALSPCS use table XDCVOLS and assume that XDCVOLS has a grouping variable named XUSEGRP. The supplied report named DCELGMGS runs on table XDCCAPD and assumes that XDCCAPD has a grouping variable named XUSEGRP. Because XUSEGRP is based on the disk volumes at your site, it is not a supplied variable. You need to define XUSEGRP for both tables before running these reports. This section will walk you through the steps to define it.

It is typical to define XUSEGRP as a formula variable. After you have the XUSEGRP formula variable defined in these tables, you can also use it in a WHERE expression as applicable. For instance, if you want to report only on system disks, you can use XUSEGRP in a local or global WHERE expression like this:

   XUSEGRP = 'SYS'

You can add formula variables to PDB tables using the IT Service Vision %CPDDUTL macro or the IT Service Vision server interactive interface. It is generally preferable to use the %CPDDUTL macro for your production PDB because the macro documents what you did exactly and enables you to add the variables to multiple tables easily and accurately.

To use the %CPDDUTL macro, you must be running IT Service Vision on the server host. Then follow these steps:

If your PROGRAM EDITOR window has pull-down menus, select Globals, Command, and Command Line.
Note: For more information about the PROGRAM EDITOR window and about pull-down menus and command lines, see Shared Appendix 2: Navigating SAS Windows.
Starting at the first line in the body of the PROGRAM EDITOR window, type in a program like one of these:
- If your disks can be grouped by volser name, you can use a program like this one, which assumes that the first three characters of the volsers are a group identifier. (Remember to change the statements to reflect your site's naming conventions.) (The #n's refer to the notes that follow.)
```
/* Add formula variable XUSEGRP to table XDCVOLS */
   %CPCAT ; CARDS4 ;                                                  #1
   SET TABLE NAME=XDCVOLS ;

      DELETE FORMULA NAME=XUSEGRP NOERROR;
      CREATE FORMULA NAME=XUSEGRP                                     #2
         LABEL='VOLUME*USAGE*GROUP'
         LENGTH=6
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            XUSEGRP = SUBSTR(DCVVOLS,1,3);                            #3
            IF XUSEGRP IN('HSM','MIS','APP','TST','SCR') THEN ;
               ELSE IF XUSEGRP IN('JES','MVS','SYS') THEN XUSEGRP='SYS';
               ELSE XUSEGRP='MSC';
                }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DISK VOLUMES'
         ;

      BUILD VIEWS NAME=XDCVOLS ;                                      #4
   ;;;;                                                               #5
   %CPCAT(CAT=SASUSER.FORMULAS.XUSEGRP.SOURCE) ;                      #6
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.XUSEGRP.SOURCE,LIST=Y);         #7
```
  1. Use the IT Service Vision %CPCAT macro to put IT Service Vision CPDDUTL control statements into a SAS catalog source member. CARDS4 indicates that the control statements continue until four semicolons are reached. For more about the %CPCAT macro, which is one method for storing CPDDUTL control statements, see the Macro Reference documentation for IT Service Vision.
    As another method, you could put the control statements in a flat file or PDS member instead of a source entry. In that case, remove the lines starting from %CPCAT to the line with four semicolons, and change the %CPDDUTL parameter from ENTRYNAM=SAS-catalog-entry-name to FILENAME='filename' and specify the fully qualified name of the flat file.
  2. %CPDDUTL control statement to create the XUSEGRP formula variable.
  3. SAS DATA step code to compute the value of the XUSEGRP formula variable based on the first three characters of the variable DCVVOLS (the volume name). In the example shown, if the first three characters are 'HSM', 'MIS', 'APP', 'TST' and 'SCR', they remain the value of XUSEGRP. If the prefix is 'JES', 'MVS', or 'SYS', XUSEGRP is set to 'SYS'. In all other cases, XUSEGRP is set to 'MSC'.
  4. Rebuilds the views for the XDCVOLS table so that they include the XUSEGRP formula variable.
  5. Terminates the CPDDUTL control statements. The four semicolons must start in column 1.
  6. Invokes the %CPCAT macro, to put all the control statements above into SASUSER.FORMULAS.XUSEGRP.SOURCE, which is a SAS catalog entry.
  7. Runs the control statements in SASUSER.FORMULAS.XUSEGRP.SOURCE.
- If your disks can be grouped by device address, use a program like this that assigns group by device address range. Change the XUSEGRP assignments as needed to represent groupings of devices at your site. (The #n's refer to the notes that follow.)
```
/* Add XUSEGRP to XDCVOLS */
   %CPCAT ; CARDS4 ;
   SET TABLE NAME=XDCVOLS ;

      DELETE FORMULA NAME=XUSEGRP NOERROR;
      CREATE FORMULA NAME=XUSEGRP
         LABEL='VOLUME*USAGE*GROUP'
         LENGTH=6
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            IF 800X LE DCVDVNU LE 815X THEN XUSEGRP='APP';            #1
               ELSE IF 900X LE DCVDVNU LE 907X THEN XUSEGRP='PROD';
               ELSE IF 0A00X LE DCVDVNU LE 0A15X THEN XUSEGRP='TEST';
               ELSE IF 0601X LE DCVDVNU LE 0605X THEN XUSEGRP='SYSDA';
               ELSE XUSEGRP = 'MISC';
                }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DISK VOLUMES'
         ;

      BUILD VIEWS NAME=XDCVOLS ;
   ;;;;
   %CPCAT(CAT=SASUSER.FORMULAS.XUSEGRP.SOURCE) ;
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.XUSEGRP.SOURCE,LIST=Y);
```
  1. This code defines XUSEGRP based on the values of DCVDVNU, the device address. If the device address is between the hex numbers 800 and 815, the value of XUSEGRP is 'APP', and so on. If the device address is not in any of the specified ranges, the value of XUSEGRP becomes 'MISC'.
- If your disks do not have a convention that allows for grouping by the methods shown above, you could use a program like the one that follows. It creates a user-written SAS format named $USEGRP in which each volser is "manually" assigned to a group. Then it stores the format in SITELIB and uses the format in the definition of the XUSEGRP formula variable. (Remember to change the literals to the ones that are appropriate at your site.)
  Note: Storing the format in SITELIB (in the PROC FORMAT step) requires update access to SITELIB. For update access to SITELIB, use the SITEACC=OLD parameter, in addition to any other %CPSTART parameters, when you invoke IT Service Vision on the server host.
  
  Note: If you want the format to apply only to one PDB, make that PDB the active PDB and then use DICTLIB instead of SITELIB. Storing the format in DICTLIB (in the PROC FORMAT step) requires update access to the PDB. For update access to the PDB, use the DISP=OLD parameter, in addition to any other %CPSTART parameters, when you invoke IT Service Vision server in batch on the server host. Or select WRITE access when you activate the PDB through the interactive interface. (The #n's refer to the notes that follow.)
```
/* CREATE A FORMAT NAMED $USEGRP AND STORE IT IN SITELIB */
   PROC FORMAT LIBRARY=SITELIB.CPFMTS;
      VALUE $USEGRP
         'VOL001','VOL003','VOL005' = 'SYS'                           #1
         'VOL111','VOL112','VOL113' = 'SYS'
         'VOL333','PAG888','PAG889' = 'SYS'
         'DISK01','DISK99'          = 'PROD'
         'ATM001','ATM002','ATM003',
            'ATM004','CIC002'       = 'CICS'
         OTHER                      = '****';
   RUN;


   %CPCAT ; CARDS4 ;
   SET TABLE NAME=XDCVOLS ;

      DELETE FORMULA NAME=XUSEGRP NOERROR;
      CREATE FORMULA NAME=XUSEGRP
         LABEL='VOLUME*USAGE*GROUP'
         LENGTH=6
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            IF DCVVOLS=' ' THEN DCVVOLS=' ';                          #2
            XUSEGRP = PUT(DCVVOLS,$USEGRP.);
            }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DISK VOLUMES'
         ;

      BUILD VIEWS NAME=XDCVOLS ;
   ;;;;
   %CPCAT(CAT=SASUSER.FORMULAS.XUSEGRP1.SOURCE) ;
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.XUSEGRP1.SOURCE,LIST=Y);
```
  where DCVVOLS is a supplied variable in the XDCVOLS table.
  1. The format definition given will output the value of 'SYS' when the input value (in the variable to which the format is applied) is 'VOL001', 'VOL003', 'VOL005', 'VOL111', and so on.
  2. This formula variable is defined using the $USEGRP format built above. XUSEGRP gives the value 'SYS' if DCVVOLS, the volume name, is 'VOL001', 'VOL003', 'VOL005', 'VOL111', and so on. If a volume is found that is not in the list that was used in building the format, then the value of XUSEGRP is the code ('****'). Results with '****' may indicate that you need to update the format.
    Note: The first assignment statement forces SAS to define DCVVOLS as a character variable. This is necessary only for the syntax check that takes place when the variable is created.
Look for the longest literal value that you assigned to the XUSEGRP variable (or the $USEGRP format). If it is longer than the value of LENGTH, increase the value of LENGTH to the number of characters in that literal.
Before submitting this program, save a copy of it for use in a later step. On your PROGRAM EDITOR window's command line, type
```
   SAVE SASUSER.SAVE.XUSEGRP1.SOURCE
```
and press ENTER or RETURN.
To submit the program, on your PROGRAM EDITOR window's command line, type SUBMIT and then press ENTER or RETURN.
To examine the effect of defining the XUSEGRP formula variable in the XDCVOLS table, follow this path from the IT Service Vision main menu:
```
   PDB Data -> Examine PDB Data ->
   XDCVOLS -> ItemActions ->
   Browse Data or View Data
```
Note: You can also browse the data by using other methods. For more information, see MVS Appendix 4: Alternate Data Browsing Method.
To copy the program back to the PROGRAM EDITOR window, on your PROGRAM EDITOR window's command line type
```
   COPY SASUSER.SAVE.XUSEGRP1.SOURCE
```
and press ENTER or RETURN.
Edit the program to change occurrences of table name XDCVOLS to table name XDCCAPD, to change occurrences of variable name DCVVOLS to variable name UCVOLSR, and to change entry name XUSEGRP1 to entry name XUSEGRP2.
Again, save a copy of it in case you need it another time. On your PROGRAM EDITOR window's command line, type
```
   SAVE SASUSER.SAVE.XUSEGRP2.SOURCE
```
and press ENTER or RETURN.
Submit the revised program. It will add formula variable XUSEGRP to table XDCCAPD.
Examine the effect of defining the XUSEGRP formula variable in the XDCCAPD table. For details, see the corresponding step for the XDCVOLS table.

Note: Using the interactive interface, you can edit the formula if you want to construct the values in a different way. For more about the construction of formula variables, follow this path from the main menu:

   Help -> Help Index ->
   formula variable table calc -> ItemActions ->
   Browse Help

Similarly, you can edit the $USEGRP format if you want to construct the values in a different way. For more about the construction of user-written formats, see Chapter 18, "The Format Procedure" in SAS Procedures Guide, Version 6, Third Edition.

For more information about the %CPDDUTL macro, follow this path from the IT Service Vision main menu:

   Help -> Help Index ->
   CPDDUTL macro dictionary utility ->
   ItemActions -> Browse Help

Note that the Help Index also has a help entry for each of the CPDDUTL control statements and for the %CPCAT source catalog utility. Also, you can see an example of the use of %CPDDUTL in the Macro Reference documentation for IT Service Vision.

Reporting (via PROJECT) on Dataset Space

You are likely to want to look at dataset space information based on some kind of grouping, like project. Because the grouping variable would be based on the projects at your site, it is not a supplied variable. You must define it based on your site's reporting needs. This section will walk you through the steps to define a grouping variable named PROJECT.

It is typical to define PROJECT as a formula variable. After you have the PROJECT formula variable defined in the XDCDSET table, you can use it in a WHERE expression when applicable. For instance, if you want to report only on the RPS project, you can specify a simple local or global WHERE expression like this:

   PROJECT = 'RPS'

   PROJECT = 'REDESIGN PRODUCTION SYSTEM'

depending on the way that you set up the PROJECT formula variable.

You can add formula variables to PDB tables using the IT Service Vision %CPDDUTL macro or the IT Service Vision interactive interface on the server host. It is generally preferable to use the %CPDDUTL macro for your production PDB because the macro documents what you did exactly and enables you to add the variables to many tables easily and accurately. The %CPDDUTL macro can be run interactively or in batch/background.

For you to use the %CPDDUTL macro interactively, you must be running IT Service Vision on the server host. Then follow these steps:

If your PROGRAM EDITOR window has pull-down menus, select Globals, Command, and Command Line.
Note: For more information about the PROGRAM EDITOR window and pull-down menus and command lines, see Shared Appendix 2: Navigating SAS Windows.
Starting at the first line in the body of the PROGRAM EDITOR window, type in a program like one of these:
- If your dataset names are constructed such that the first n characters are a project code, you may want to create the PROJECT formula variable by using a program like this (in this example, n is 3):
```
/* Add PROJECT to XDCDSET */
   %CPCAT ; CARDS4 ;
   SET TABLE NAME=XDCDSET ;

      DELETE FORMULA NAME=PROJECT NOERROR;
      CREATE FORMULA NAME=PROJECT
         LABEL='Project Group'
         LENGTH=3
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            PROJECT = SUBSTR(DCDDSNA,1,3); }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DATASETS'
         ;

      BUILD VIEWS NAME=XDCDSET ;
   ;;;;
   %CPCAT(CAT=SASUSER.FORMULAS.PROJECT.SOURCE) ;
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.PROJECT.SOURCE,LIST=Y);
```
  where DCDDSNA is a supplied variable in the XDCDSET table. Each value of the DCDDSNA variable is a dataset name.
  
  This example results in values of PROJECT that are three-character project codes.
- If you want to translate the short code for a project into a longer label, you may want to create a format named $CPPROJ (with the appropriate literals for your site), store it in SITELIB, and then use the format to define the formula variable named PROJECT.
  Note: Storing the format in SITELIB (in the PROC FORMAT step) requires update access to SITELIB. For update access to SITELIB, use the SITELIB= parameter in addition to any other %CPSTART parameters when you invoke IT Service Vision on the server host.
```
/* CREATE A FORMAT NAMED $CPPROJ AND STORE IT IN SITELIB */
   PROC FORMAT LIBRARY=SITELIB.CPFMTS;
      VALUE $CPPROJ
         'IDC' = 'DATA CENTER'
         'MKT' = 'MARKETING ANALYSIS'
         'HRS' = 'HUMAN RESOURCES SYSTEM'
         'RPS' = 'REDESIGN PRODUCTION SYSTEM'
         ;
   RUN;

/* Add PROJECT to XDCDSET */
   %CPCAT ; CARDS4 ;
   SET TABLE NAME=XDCDSET ;

      DELETE FORMULA NAME=PROJECT NOERROR;
      CREATE FORMULA NAME=PROJECT
         LABEL='Project Group'
         LENGTH=26
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            TEMP = SUBSTR(DCDDSNA,1,3);
            PROJECT = PUT(TEMP,$CPPROJ.); }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DATASETS'
         ;

      BUILD VIEWS NAME=XDCDSET ;
   ;;;;
   %CPCAT(CAT=SASUSER.FORMULAS.XUSEGRP.SOURCE) ;
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.XUSEGRP.SOURCE,LIST=Y);
```
  where DCDDSNA is a supplied variable in the XDCDSET table. The DCDDSNA variable contains the dataset names.
  
  The value of LENGTH is set to the longest value of PROJECT.
- If at your site the dataset high-level qualifiers do not contain a code to identify the project, you may want to create a format from a separate SAS data set that contains employee information entered manually or read from an external file. Below is a simple example. If some high-level qualifiers do not always belong to the same project or if assignments of high-level qualifiers to projects change over time (for example, if people change departments but not userids), then this approach will be too simplistic.
  For example, suppose you have a flat file that looks like this:
```
   SYS1     SYSTEM
   PROD     PROD
   UID123   DEV
   UID456   TST
```
  and has information on the following variables:
```
   high-level qualifier
   project code
```
  You could write a SAS program to read the values from the flat file and write them to a SAS data set. Here is a very simple version that reads the records in the flat file shown above and writes the data to a SAS data set named SASDATA.HLQS (you must have previously allocated a SASDATA library):
```
   DATA SASDATA.HLQS;
      INFILE name-of-your-flat-file ;
      INPUT @1   HLQ     $CHAR8.
            @10  PROJ    $CHAR6.
            ;
   RUN ;
```
  You could then run that program in front of the program below.
  
  Note: Storing the format in SITELIB (in the PROC FORMAT step) requires update access to SITELIB. For update access to SITELIB, use the SITEACC=OLD parameter when you invoke IT Service Vision on the server host. (The #n's refer to the notes that follow.)
```
   /* CREATE A SAS DATA SET THAT CONTAINS A FORMAT NAME, HIGH-LEVEL
   QUALIFIER, AND PROJECT CODE IN EACH OBSERVATION. */
        DATA FMTCNTL (KEEP = FMTNAME START LABEL);                    #1
           SET SASDATA.HLQ;
           RETAIN FMTNAME  '$CPPROJ';
           START = HLQ;
           LABEL = PROJ;
        RUN;

   /* USE THE FMTCNTL DATA SET TO CREATE THE FORMAT AS SPECIFIED IN THE
   PREVIOUS DATA STEP */
        PROC FORMAT CNTLIN=FMTCNTL LIBRARY=SITELIB.CPFMTS;            #2
        RUN;

   /* ADD VARIABLE PROJECT TO TABLE XDCSET USING THE $CPPROJ FORMAT. */
   %CPCAT ; CARDS4 ;
   SET TABLE NAME=XDCDSET ;

      DELETE FORMULA NAME=PROJECT NOERROR;
      CREATE FORMULA NAME=PROJECT
         LABEL='Project Group'
         LENGTH=6
         TYPE=CHARACTER
         INTERPRET=STRING
         SOURCE={
            HLQ = SCAN(DCDDSNA,1,'.');                                #3
            PROJECT = PUT(HLQ,$CPPROJ.);
            IF PROJECT = ' ' THEN PROJECT = HLQ;
            }
         LEVELS='DETAIL DAY WEEK MONTH YEAR'
         KEPT=YES
         SUBJECT=''
         DESCRIPTION='GROUPING VARIABLE FOR DATASETS'
         ;

      BUILD VIEWS NAME=XDCDSET ;
   ;;;;                                                               #4
   %CPCAT(CAT=SASUSER.FORMULAS.XUSEGRP.SOURCE) ;
   %CPDDUTL(ENTRYNAM=SASUSER.FORMULAS.XUSEGRP.SOURCE,LIST=Y);
```
  where DCDDSNA is a supplied variable in table XDCDSET. The DCDDSNA variable contains dataset names.
  1. The first step generates a SAS data set with one observation per high-level-qualifier, and each observation has these variables: FMTNAME, which is the constant $CPPROJ; START, which is the high-level qualifier; and LABEL, which is the project name.
  2. The second step reads in the data set from the first step and uses that as an input control data set. Based on that, PROC FORMAT creates a user format named $CPPROJ and stores it in SITELIB.CPFMTS.
  3. The third step calculates the value of PROJECT from a supplied variable named DCDDSNA and adds this formula variable, using the format named $CPPROJ, to the XDCDSET table, much the same as previous examples did. LENGTH is set to the longest value of PROJ.
  4. The four semicolons must begin in column 1.
To submit the program, at your PROGRAM EDITOR window's command line, type SUBMIT and then press ENTER or RETURN.
To examine the effect of defining the formula variable, follow this path from the IT Service Vision main menu:
```
   PDB Data -> Examine PDB Data ->
   XDCDSET -> ItemActions ->
   Browse Data or View Data
```
You can use the menu items under the View pull-down menu to view the PROJECT values for each observation. (You can also use other methods to browse the data. For more information, see MVS Appendix 4: Alternate Data Browsing Method.)

Note: If you have write access to the PDB, you can edit the formula if you want to construct the values in a different way. For more information about the construction of formula variables, follow this path from the main menu:

   Help -> Help Index ->
   formula variable table calc -> ItemActions ->
   Browse Help

Similarly, you can edit the $CPPROJ format if you want to construct the values in a different way. For more information about the construction of user-written formats, see Chapter 18, "The FORMAT Procedure" in SAS Procedures Guide, Version 6, Third Edition.

For more information about the %CPDDUTL macro, follow this path from the IT Service Vision main menu:

   Help -> Help Index ->
   CPDDUTL macro dictionary utility ->
   ItemActions -> Browse Help

Note that the Help Index also has a help entry for each of the CPDDUTL control statements and for the %CPCAT source catalog utility. Also, you can see an example of the use of %CPDDUTL in the section "Example of %CPDDUTL" in the Macro Reference documentation for IT Service Vision.