The DeskTop Reporter for IT Service Vision for MVS contains a set of reports that can be examined by running them from the supplied sample data.

Follow the instructions on the install document that is specifically for the DeskTop Reporter. When this has been completed, you should be able to log onto the mainframe. This will allow you to download your summarized data to the DeskTop Reporter.

The reports can also be run using your own data after you have run the appropriate macros on the mainframe to subset your data and prepare it for downloading. The sample program to accomplish this task can be found in the cpmisc library member cmeisdwn.

The first time, just run the job exactly as is without modifying any files. After you have run the CPEISSUM macros and created a download library, you can bring up the DeskTop Reporter on your PC.

(If you have customized the reduction levels of your PDB, you will need to make changes before you can run the existing reports. Follow the instructions for customizing the summary code later in this document.)

Download the data from MVS using SAS/CONNECT and the Download Data button. You will be prompted for the name of the SAS library that contains your downloaded data. This is the library in the download LIBNAME statement in your batch program. You will also be prompted for the name of your MXG format library. After the first time, these names are "remembered" for other downloads.

This is the default support provided with the DeskTop Reporter. Other methods are available and discussed under the help button of the download menu. The names of the datasets must remain the same to use the reports as they are currently written.

The DeskTop Reporter can also be customized to be specific for your site.

The following steps outline one way to proceed through the steps to customize the DeskTop Reporter for use at your site. First, compare the hierarchy chart shown below with the DeskTop Reporter on your PC. On the chart below, mark out any reports and/or any entire block of reports that you do not want to use at all. Now mark with a C any reports that you want to keep but to which changes need to be made. New titles, new subgroups, and different variables are examples of changes that you might want to make.

The Hierarchy of IT Service Vision for the MVS DeskTop Reporter

IT Service Vision for MVS DeskTop Reporter

• Download Data
  • Signon to MVS
  • Download Data
  • Disconnect
  • Help
  • Exit
    
    
• CPU Usage
  • CPU busy (Prime Shift)
  • Various Metrics by CPU
  • Address Space Analysis
  • Workload Analysis
  • Logical Percent CPU Busy for LPARs
  • Logical Overhead per LPAR
  • Goal Mode Metrics
    
    
• Storage Usage
  • Storage Map (Prime Shift)
  • Storage Analysis TOTAL - USED by DATE
  • Cache Analysis
    
    
• I/O Reports
  • IO Analysis by Device and Time
  • Paging Analysis by Workload
    
    
• Networks
  • NetSpy Line Report
  • NetSpy Response Times
    
    
• CICS Reports
  • Various CICS Metrics
    
    
• DB2 Reports
  • DB2 Accounting Report
  • DB2 Statistics Report
  • DB2 (More) Statistics Report
    
    
• Utilities
  • Run SAS/INSIGHT
  • Download/Print Graphs from MVS
  • Download MXG formats from MVS
    
    
• Help
  • What’s Here
  • Using It
  • Changing It
    
    
• Exit

To customize the DeskTop Reporter and still be able to incorporate any new reports that SAS might add to the Reporter, copy all the reports that you want to use into a separate application database. One is built for you in CPUSER, the IT Service Vision for MVS application databbase for user reports. You will change your SAS autoexec to point to your copy of the application.

Customizing the Menus

First copy the main menu to your application data base. At this point, you can decide whether you want to continue to point to the IT Service Vision main menu or start at the MVS DeskTop Reporter main menu. Since the MVS piece is the only one shipped in this initial release, you can change your autoexec to point to CPEEIS.CPUSER.MAIN.BLOCK and just copy this menu.

You will need to run SAS/EIS on the CPEEIS library and the CPEEIS application database. To bring up SAS/EIS, exit the DeskTop application and select Globals -> Develop -> EIS Application Builder at the top of the SAS PROGRAM EDITOR window.

Click on Build EIS.

Select the path to the CPEEIS library and select CPEEIS as the Application Database.

Select MAIN IT Service Vision for MVS DeskTop Reporter BLOCK

Then select the COPY button. Under copy to change the Database by clicking on the down arrow. Select IT Service Vision for MVS app db for user reports CPUSER. Enter the menu name, the same as in the other catalog MAIN. Select OK.

Now from the Build EIS menu, click on the down arrow under Application Database and Select CPUSER. Now you will see:

First Select USER1 and then select the Delete button at the bottom of the window. Select OK when asked to verify the delete. Now select MAIN and then the Edit button at the bottom of the window. Now you will see a Block Menu edit screen. You can change the Description and/or the Block title to a title for your organization, like Performance Reports for XYZ Corporation.

Now select Edit menu text:

Block 1 is the block menu for the Download Data button. If you are using SAS/CONNECT to download your data to the PC, you will probably want to leave this as is. Scroll down, using the scroll bar on the right, to Block 2.

Block 2 is the block menu for CPU Usage. If you don’t want ANY of the CPU reports and want to remove the block menu, select the Delete button at the bottom of the window; otherwise continue going through the menu selections, deleting any entire blocks that you don’t want to have appear on your menu. (You’ll delete individual reports during the next step.)

After you have finished customizing the main menu, select OK at the bottom of the window. You can test your main menu by selecting it in the list and selecting the Test button. After you complete your test select OK again.

Now once again, change back to the CPEEIS Application database by selecting the down arrow under Application Database: CPUSER. Select CPEEIS. Now you will see the list of objects as distributed. Scroll down in the list until you find the LISTMENU section.

Now you will want to copy all those LISTMENUs that you want to continue to use in your customized DeskTop Reporter. You’ll see this list:

To copy them into your application database, select the LISTMENU that you want to copy in your object list (like the one above.) Then select COPY from the blocks below the list. From the Copy Screen, change Database to CPUSER. Enter the name to be the same as the original name. You can change the Description, if you want. Click on OK. Continue this for each of the LISTMENUs that you want to include.

When you complete the copies, change the Application Database on the Build EIS menu to CPUSER. Now select one of the LISTMENU’s. Then click on the Edit button. From the List Menu window, click on Edit menu text. This window shows the report Description; the description is what you see in the list of reports when you select this menu. At the top of the screen you see :

Library:

Database: CPEEIS

Name:

Type:

This will provide the complete path name for the report that will be run when you select this menu item.

If this report is to be deleted, just click on the Delete button at the bottom of the window.

If there are NO changes to be made to this report, you can leave the path as it is now. But, if you marked this report with a C in your hierarchy chart, you will want to change the database name from CPEEIS to CPUSER. You will copy this report into your database and make changes in a later step.

Continue through the menu items by using the scroll bar on the right of the screen. When you have completed the modifications for this menu, select the OK button at the bottom of the screen. Then select OK again.

Continue to modify the remaining LISTMENUs in this same way.

When you have completed the modification of the LISTMENUs, change your application database back to CPEEIS from the Build EIS screen.

Customizing Individual Reports

Now by using the table shown above (which reorders the object list in the same order as the report hierarchy), select the reports (from the list on the Build EIS screen) that you marked with a C on your original hierarchy chart. Copy these to CPUSER. After the copy has been completed, from the instructions given below follow the directions for the report type that you are modifying.

To copy the report, select the report in the list. Then select the Copy button at the bottom of the screen. Change your Database to CPUSER. You can keep the name of the report the same or change the name. The choice should depend on how you want to incorporate any future report changes that SAS makes to the distributed DeskTop Reporter. If you keep the name the same as the original report, you will be able to match your report with any modified report.

If you are making changes that require changes to the format of the data being downloaded from the IT Service Vision for MVS PDB, you will need to make those changes first, download your data to the PC, update the MetaBase in EIS on the PC, and then update the report.

So let's look at how to make the changes to the CPEISSUM macro on MVS. Using the chart given below, you will be able to determine which of the cpeissum macros given in the cmeisdwn jcl in the CPMISC library is used to download data for which reports.

If you have removed reports, you can also remove the macros that download the data for those reports. The only exception is the Storage reports that include data from RMF 71 and 72 records. Those records must both be included in the same macro in the correct 71 and then 72 in order for that report to run correctly.

Customizing data downloaded from MVS

Subsetting from PDB using CPEISSUM

Allocate the PDB and set the IT Service Vision system options using the %CPSTART macro.

Run the %CPEISSUM macros on MVS. See an example job in member CMEISDWN in the CPMISC PDS on MVS that was installed with Release TS450 of IT Service Vision for MVS.

It contains:

libname download ‘your.download’;

This is a SAS data library that has already been allocated by you, either from the SAS display manager or by using ISPF to allocate it using the appropriate file characteristics, which are RECFM=FS with half track blocking.

If you enter the LIBNAME statement from SAS display manager, SAS will allocate the library for you with the appropriate characteristics.

Remember this name because you will enter it from your PC when you are prompted for the download data set name from the Download Data button of the DeskTop Reporter.

The %CPEXCEPT ... statement can be removed.

The following %CPEISSUM macros are given:

--------------

The first invocation of %CPEISSUM asks for 1 day at the DAYS level of the PDB to be downloaded for table XDBACCT. The intermediate data set will be loaded into WORK.XDBACCTD because outlib=work (the D is added to the table name because it's day level). This macro invocation is using the self-summarization mode, so the source code in PGMLIB.CPSAMP.XDBACCTD.SOURCE will be run after the data is subset. This data is used for the report xxxx as shown in the table above.

%CPEISSUM(XDBACCT,

n_detail=0,

n_days=1,

n_weeks=0,

n_years=0,

outlib=work);

-----------------------

This next invocation of the %CPEISSUM macro asks for 2 days of DAY level data from the PDB for the 3 tables XDBSTA0, XDBSTA1 and XCITRAN. Again the extracted data will be loaded into WORK.XDBSTA0D, WORK.XDBSTA1D and WORK.XCITRAND. And the source members in PGMLIB.CPSAMP.XDBSTA0D.SOURCE, PGMLIB.CPSAMP.XDBSTA1D.SOURCE, and PGMLIB.CPSAMP.XCITRAND.SOURCE will be run. This data is for the reports xxx,xxx and xxx as shown in the table above.

%CPEISSUM(XDBSTA0 XDBSTA1 XCITRAN,

n_detail=0,

n_days=2,

n_weeks=0,

n_years=0,

outlib=work);

------------------------

The following invocation of the %CPEISSUM macro downloads data from DCOLLECT. However, no reports have been done using this data. You may, of course, build your own DeskTop Reports using SAS/EIS on the PC.

%CPEISSUM(XDCVOLS XDCCAPT XDCCAPD,

N_detail=1,

n_days=1,

n_weeks=0,

n_years=0,

outlib=work);

------------------------

The following macro extracts data from two different levels of the PDB. Two days are extracted from DETAIL and 7 days are extracted from the DAY level of the PDB. The source code to run against the detail level is in PGMLIB.CPSAMP.XTY70.SOURCE and PGMLIB.CPSAMP.XTY72.SOURCE. The data from the detail level will be stored in WORK.XTY70 and WORK.XTY72, and the data from the DAY level will be stored in WORK.XTY70D and WORK.XTY72D. See the table above for the reports this data is used in.

%CPEISSUM(XTY70 XTY72,

N_detail=2,

n_days=7,

n_weeks=0,

n_years=0,

outlib=work);

------------------------

The following macro extracts data from the tables named at the detail level (2 days) and from the day level (1 day). The extracted data is stored in WORK and the source code will be found in SOURCE members in PGMLIB.CPSAMP. The XCATY data does not have a report written to use it.

%CPEISSUM(XTY71 XTY74 XASM70P XNSLINE XNSLU XCA90 XCATY,

N_detail=2,

n_days=1,

n_weeks=0,

n_years=0,

outlib=work);

------------------------

This last macro invocation extracts 1 day of detail data, stores it in WORK.XTY72GOD and runs the source code in PGMLIB.CPSAMP.XTY72GOD.SOURCE.

%CPEISSUM(XTY72GO,

N_detail=0,

n_days=1,

n_weeks=0,

n_years=0,

outlib=work);

---------------------

Customizing the Summary Code

Copy the source members that are to be changed from PGMLIB.CPSAMP into ADMIN.CPSAMP. That way you have a backup.

Add source members in ADMIN.CPSAMP or another library of your choice to summarize the data for any reports that need to be summarized. The name of the library chosen is coded as a parameter of the CPEISSUM macro. Example:

%CPEISSUM(XTY70,sourcat=ADMIN.CPSAMP....

The name of the SOURCE member used is named the same as the table. The DAY level has a D on the end, the WEEK level, a W, etc. Examples of the SOURCE members you need to build are in PGMLIB.CPSAMP.tablename.SOURCE. All IT Service Vision table names start with an X.

To change source members from SAS Display Manager, use the copy command from the command line to copy the source member into the program editor.

• COPY PGMLIB.CPSAMP.tablename.SOURCE
• Then you can save it into ADMIN by using this statement from the command line.
• SAVE ADMIN.CPSAMP.tablename.SOURCE
• Hint: You can use a ? to recall previous commands from the command line.

The table below gives you a list of source members in pgmlib.cpsamp in the same order as your hierarchy table.

Customizing List of Data that is Downloaded

You can then modify the list of data to be downloaded in CPEEIS.DOWNLOAD to remove those data sets that no longer need to be downloaded.

Use either SAS/FSP or data step code to modify this list according to the table given below.

To use FSEDIT, enter the following command from the command line:

fsedit cpeeis.download

If you are not licensed for SAS/FSP (on which FSEDIT is dependent), you can use a DATA step to update the entries. To see what is currently in the data set, run the following code from the PROGRAM EDITOR window:

proc print data=cpeeis.download; run;

The data set contains only three variables:

rptname - 'name of the report in the desktop application';
hostname - 'name of data in the download dataset on the host';
dsktname - 'name of data on the PC. name will be defined in the
Metabase and in the report object'

If you want to add datasets to the download function, create a dataset called CPEEIS.USERADDS, rather than adding them to the supplied CPEEIS.DOWNLOAD dataset.

You can create the dataset with code like this:

data cpeeis.useradds;
input rptname $ hostname $ dsktname $;
cards;
RNAME1 HNAME1 DNAME1
RNAME2 HNAME2 DNAME2
RNAME3 HNAME3 DNAME3
run;

If, in the future, you want to add even more datasets, you can modify your version of the above code to add your new datasets. This, however, implies that you would be storing a copy of that code. Alternatively, you could use the following code to add observations to an existing CPEEIS.USERADDS dataset.

data newones;
input rptname $ hostname $ dsktname $;
cards;
RNAME4 HNAME4 DNAME4
RNAME5 HNAME5 DNAME5
RNAME6 HNAME6 DNAME6
run;
proc append base=cpeeis.useradds data=newones;
run;

If you want to stop downloading some summary data that is listed in the supplied CPEEIS.DOWNLOAD dataset, currently you will have to modify that dataset to remove it.

The report name can be anything for the time being, but will be replaced by the name of the new report you build.

All datasets that are specified in both CPEEIS.USERADDS and CPEEIS.DOWNLOAD

are downloaded when you select the Download Data icon.

Download your Data

Next you will need to download your data to the DeskTop. The download can be done using the menu if you first modify the list of datasets to be downloaded. See instructions above.

Defining Data to the MetaBase

After the data has been downloaded, you will define your data to the MetaBase.

From Globals, select Develop -> EIS Application Builder. Then select Metabase.

In the CPEEIS library, a CPUMETA metabase has been built for your use. Add your table to the metabase and customize the attributes for the table and columns as appropriate. You can also copy an existing definition from the supplied MetaBase to your User MetaBase by selecting the table definition to copy, selecting the Copy Block at the bottom of the window. Change the database location to CPUMETA and keep the name of the table the same as that copied. See the instructions under each individual report type for how to customize the MetaBase for that report type.

Customizing Individual Report Types

After the metabase has been built, you can build and/or modify a report.

From the Main Menu select Build EIS. In the CPEEIS library, an application database for user extensions has been provided. Select CPUSER to add your reports.

If you build a new report, don’t forget to modify your copy of the menu to access your report.

The Workload Analysis Report using a Notepad to list the performance groups at the site.

To customize the performance group list in CPEEIS.EXAMP.PGDESC.SOURCE. Enter the following command and make any changes you need:

NOTEPAD CPEEIS.EXAMP.PGDESC.SOURCE

In the next release of the DeskTop, a format will be used to label the performance groups appropriately and the notepad will be used for other instructions.

MULTCOL

• Source Code Structure

  Uses a PROC SUMMARY.

  Weight duration used for interval tables.

  Var statement includes all metrics for inclusion in the table at the bottom.

  Class statement defines class variables in REVERSE ORDER to that desired in the drill hierarchy in the report. In other words, if you want the report to begin by showing your 3 machines, when you select a machine, you will see that machine broken down by shift and when you select a shift, you will see an hourly breakdown. Hierarchy will be MACHINE SHIFT HOUR. Then the CLASS statement will be: class hour shift machine;

• MetaBase Structure

  After adding the table to your MetaBase, select the table to be modified. Select the Add button beneath Attributes. Select HIERARCH. Define your drill-down hierarchy. Select SUMMARY since the data has already been summarized using PROC SUMMARY. The two lists should match.

  Then select Columns beneath Tables. This will show a list of metrics to be reported on. For each metric in the report, select that metric and the the ANALYSIS metric.

• Building the Report Object

  From the Object List, select the Add button (or Edit). Select Business Reports -> Multi-column report. Select the Build button to the right.

  Enter (or Edit) the report name and description. Select the arrow beside Data Set. You will be prompted with all data sets defined to the MetaBase appropriately for this type of report. Select Start Drill and select the first variable in your hierarchy list. Select the Variables arrow and select a list of variables. Select Statistic, if appropriate. From Customize, you can select chart type, colors, labels and formats and totals. Use the Test button to test the object definition. When satisfied, select OK.

GROUPBAR

• Source Code Structure

  Uses PROC SUMMARY.

  May need to reformat the data if there is not an appropriate grouping variable, like performance group or lpar. See XASM70P for an example.

  Var statement lists metrics that may be reported in the bar chart.

  Weight duration is used for interval tables.

  Class statement is again in reverse drill order.

• MetaBase Structure

  Define the data set with a HIERARCHY and with the SUMMARY attribute. Define those variables to be shown on the bar as ANALYSIS variables.

• Building the Report Object

  Select Add (or Edit) for the report. Then enter name and description. Select the data set from the list. Select the Start of drill from the hierarchy list. Select the variables to be shown on the bar. Use the Test Button.

  Help from EIS: Select Tutorial -> Objects -> Object Groups -> Business Graphs -> Grouped Bar Chart CHART

• Source Code Structure

  Same format at GROUPBAR. See XTY70, xty70wkls.

• MetaBase Structure

  Define the dataset with a HIERARCHY and with a CATEGORY variable and an ANALYSIS. The CATEGORY variable will define the bars on the chart. The ANALYSIS variable will determine the height of the bars. The hierarchy can also be used to define the columns.

• Building the Report Object

  Define the name and the description. Select the data set from the list given when the button is clicked. Select the variable from the list given. Select the Type of Chart. Select Customize -> Statistical Options to select the Analysis Variable and Statistic. Select Bar Chart subgroups to select any subgroups on the chart.

  Help from EIS: Select Tutorial -> Objects -> Object Groups -> Business Graphs -> Charts EXPANDER

• Source Code Structure

  Uses PROC SUMMARY.

  Var defines variables to appear on columns of the report.

  Class defines levels that will be expanded

  Weight duration for interval tables.

• MetaBase Structure

  Must have a HIERARCHY attribute and an ANALYSIS variable. Also define the SUMMARY attribute since you used PROC SUMMARY.

• Building the Report Object

  Define the name and the description. Select the data set. Start drill variable from the hierarchy and the analysis variables. Press Test to see if the report is as you want it.

  Help from EIS: Select Tutorial -> Objects -> Object Groups -> Business Reports -> Expanding Report

In Summary…

This is an experimental release of the DeskTop Reporter.

Tech Support may not be able to answer questions, so you may be referred to the IT Service Vision for MVS Marketing and/or Development Staff for more information. However, to allow us to track calls more closely, please call Tech Support first.

CPEISSUM Macro Reference:

The CPEISSUM macro contains several modes of operation:

Auto-summarizing with PROC SUMMARY
Auto-summarizing with PROC MDDB
Self-summarizing

The following parameters are used for both self- and auto-summarizing modes:

tablist	List of table names to summarize for EIS.
Outlib=WORK	Library to subset/summarize data into
outmems	Names of resulting summarized data sets or mddbs to which blank, D, W, M, Y is appended for redlevels DETAIL, DAY, WEEK, MONTH and YEAR. Names must be 7 or fewer characters long and separated by blanks. Default value is tablenm (from tablist) with blank, D, W. M, or Y appended to it for DETAIL, DAY, WEEK, MONTH or YEAR. Not e that this also controls the name of the .source entry to in included in the self-summary mode.
Outlabs=	The labels to be associated with outmems=, separated by !
n_detail=0	Number of days of detail data to extract.
N_days=0	Number of days of day data to extract.
N_weeks=0	Number of weeks of week data to extract.
N_months=0	Number of months of month data to extract.
N_years=0	Number of years of year data to extract.
Where=	expression used to subset the data before summarization. Only those observations for which the expression evaluates as true will be part of the summarized data. A typical value for this might be WHERE=SHIFT=1 to restrict analysis to the prime shift data.
Internal=	Internal use only

The following parameter is for self-summarizing mode only:

sourcat=PGMLIB.CPSAMP

Catalog containing source entries named for the table being extracted.

The following parameters are for auto-summarizing mode only:

var=	List of vars to calculate summary statistics on. The presence of this list causes CPEISSUM to summarize the data using the SUMPROC= procedure.
Class=	List of class variables to summarize on.
Sumproc=MDDB	Procedure to use to summarize data (SUMMARY)
hiers=	Lists of variables to make up named hierarchies for summproc=mddb only. Multiple hierarchies are separated by a vertical bar (|). Hierarchies are named by the variable names compressed (blanks removed). For example, hiers=machine Data would produce the statement hierarchy Machine Date/ name=MachineDate display=yes; for PROC MDDB. The named hierarchy can then be used directly in EIS report applications.
Stackvar=NO	If YES, the variables listed in var= are assumed to represent the parts of another metric that you ultimately want to produce a stacked bar chart of. This options typically appears with the outmems= parm to control the name of the resulting summarized data set. The only metrics in the resulting summarized data set are the class= variables plus 2 new variables--one numeric var ("_STKVALU") that contains the value of one of the variables to be stacked and one character var ("_STKTYPE") that contains the name of the var whose value is in _STKVALU.