Phase 1: Decide Implementation Strategy

The implementation of Control-D at your site will have a major impact on your current distribution system. You have the potential to revolutionize the distribution methods currently used. Few projects that you undertake will have such a direct impact on the end user environment.

Only in the past few years have computer departments adapted themselves into a true service industry. Previously, the services supplied to users were controlled by the computer department. Now that many end users have become more computer literate, they are pushing the computer department to supply the services that they require for their business needs. To measure the level of services supplied, many computer departments have introduced Service Level Agreements that contract them to supply defined levels of service.

Data input methods have come a long way since the days of punched cards and paper tape. The end users now have user friendly online facilities for the input of data to the computer systems. However, the output service supplied to them has changed very little, if at all. One of the few enhancements was the introduction of laser printing, allowing users to receive slightly less bulky reports in a nicer format.

With Control-D, you have the power to bring the end users’ output service up to the same standard that they expect from their current input facilities. (If you asked users to revert to filling in forms and inputting data in punch card format, they would be extremely unhappy.) The same level of enhancement, and more, will be achieved when you implement Control-D in your production environment. The output management service you supply will provide the users with a new dimension of productivity.

Because the output service has changed very little over the last twenty years or so, some users may be apprehensive about the introduction of new facilities. This guide takes into consideration such end user apprehensions in its approach to implementation. After all, it is to the end user that we provide service. It is our responsibility not to scare the end user with the thought of change, but to show them the benefits that they will gain.

Some of the questions we will be answering in this phase are:

How do I measure the success of the project?
What do I want to achieve from the implementation?
What are my priorities for implementation?
What should I do first?
What resources do I need, and in what quantity?

Inputs

None

Outputs

At the end of this phase you will have:

Performed a distribution system review
Decided your project objectives
Decided your implementation strategy
Selected a pilot application
Selected the super users
Assigned resources to the project.

Review the Current System

To help you set your priorities for Control-D implementation, and to enable you to measure the success of your Control-D implementation, you must first know the status of your current system. Only by analyzing and measuring your current system will you be able to gauge how successful your implementation has been.

In this section, we will outline what information you should gather before you start the implementation.

How Much Output Do You Currently Produce?

One of the most common objectives of implementing Control-D is to reduce the amount of printed output. To help you assess the savings you can achieve by implementing Control-D, you need to find out how much output you currently print. You should produce figures that reflect all the printing you do. Perhaps at specific periods you create large volumes of output. If you do, then ensure that these figures are included in your statistics. You can use SMF Printer Statistics to gain precise values.

Suggestion

The most obvious way to reduce printed output is to move reports to an Online Viewing environment (described in Phase 9: Online Viewing Implementation) and to prevent unwanted data from being printed (see "How Much Unwanted Information Is Sent to Users" later in this phase). Once the implementation has started, one of your tangible goals might be to reduce the number of pages printed. Be aware that the introduction of new applications or natural increase in growth will distort your figures.

Information Required

Table 1 Output You Currently Produce

Item	Estimate
Daily average number of pages printed:
Weekly average number of pages printed:
Monthly average number of pages printed:
Quarterly average number of pages printed:
Yearly average number of pages printed:

How Big is the Spool and How is it Used?

The spool was intended to be a transient area for report printing and this is exactly how Control-D uses it. Implementing the CDAM Direct Write feature (described in Phase 6: Implement CDAM Direct Write) will mean that you no longer overload the spool at peak periods. Typically, you will be able to reduce the size of the spool to about one third of its original size, giving you back some valuable DASD space. Another hidden benefit of the CDAM Direct Write feature is that you will see a decrease of about ten percent in the execution time of batch jobs creating large report outputs.

The spool size for Site X is three single density 3,380s (2,655 cylinders). The utilization of spool normally averages between forty and fifty percent. However, at end of month, large volumes of output are created and spool utilization reaches around ninety percent. Once the monthly reports are printed, the spool returns to its average utilization. The size is set to cope with the peak period so that the spool utilization does not reach 100% and JES stops processing.

Suggestion

In the above example, using the CDAM Direct Write feature, the user would retrieve two of the 3380s: one to be used for the CDAM datasets, and one for any other purpose. We achieve the reduction in spool size by writing output directly into a compressed format on DASD. The compression rate is between 30% to 70% depending on the type of data being compressed. Therefore, in the above example, we would have one volume allocated for spool, one for the CDAM datasets, and one returned for other purposes.

If one of your objectives is to reduce the size of the spool area and to retrieve some DASD space, you should implement CDAM Direct Write to your batch jobs. For now, find out what the minimum, average and peak utilizations are, and how much space is allocated to the spool area.

Information Required

Table 2 Spool Size and Usage

Item	Estimate
Size of spool area (cylinder):
Minimum % utilization:
Average % utilization:
Maximum % utilization:

How Much Unwanted Information is Sent to Users?

Redundant data is inherent in most distribution systems. Many reports sent to users contain irrelevant or useless information. Likely candidates for redundant information are duplicated reports.

Ten full copies of a Sales Analysis report are printed each day using the JCL COPIES parameter. The report contains details of sales by region followed by a summary of sales by region. Each region is only interested in detail figures from its own area and summary information for all areas. In this environment, there is no way to exclude the unwanted detail data, so each region receives a full copy of the report on a daily basis.

We may also want to exclude data depending on the day of the week. Perhaps a specific user receives a General Ledger report on a daily basis (Monday – Friday), but the only day the user actually wants the full report is Friday; on the other days, the summary section would suffice.

Suggestion

You can reduce the amount of printed output by eliminating any redundant data dispatched to users. We call this "report pruning" and we use the Control-D decollating missions (described in Phase 3: Design Decollating Missions) to distribute exactly the parts of a report that each user requires. We will be asking the users to identify redundant data that we can exclude when we undertake a user survey in Phase 8: Production Implementation.

In What Quantities Are Resources Required for Report Backups?

One of your objectives may be to implement a secure archiving facility for your reports (described in Phase 5: Implementation of System Administration Tasks) so that you can restore them easily if they are lost or destroyed. If you currently back up reports, find out in what quantities resources are required. This information will be useful for comparison after you implement Control-D in production.

Site X creates reports to production print files from batch JCL steps. They then use IEBGENER to backup and print the production files. For example:

Create PROD print files from the application program.
Use IEBGENER to create BKUP version of report files.
Use IEBGENER to print PROD print file to spool.

In this example, the user may have two versions of the same report file on DASD at any one time.

Suggestion

Using Control-D, you need only create the file directly to spool and Control-D will automatically create a compressed version of the file when Control-D processes it. (You can, of course, use the CDAM Direct Write option to bypass the write to spool step altogether.) The compressed datasets will give you between 30% to 70% reduction of allocated space. We recommend that all reports be backed up using Control-D’s archive facility (described in Phase 5: Implementation of System Administration Tasks), so that you can easily restore reports if required. Control-D will use your current backup product to perform archive and restore functions. The supported products are:

FDR/ABR
DF/DSS
DF/HSM
DMS/OS
ASM2
ARCS
In-House Application

Information Required

Table 3 Quantities of Resources Required for Backup

Item	Estimate
DASD space for report files:
Tape and/or cartridge pool size:

What is the Current Level of Report Reruns?

Some of the savings you will achieve after Control-D has been implemented may not be immediately apparent. Recreation of a lost or destroyed report may seldom be required, but when it is performed, it consumes a large number of resources.

The user receives a 1,000-page report of which two important pages are unreadable. The user contacts the Help Desk or Support Section requesting the reproduction of the two pages. Once the specific report is identified, which may take some time, the Help Desk opens a problem report, and issues a report rerun request. They either have to regenerate the report at program level or recreate the report from a backup file, both of which are resource intensive. They cannot reproduce just the two unusable pages, so they recreate the entire 1,000-page report. Sometimes the print operators try to forward space to the required pages; more often the entire report is reprinted and distributed. The personnel resources taken up are difficult to quantify, and so is the impact on the business itself. The resources used are as follows:

Problem and/or change management resources.
Personnel resources to set up rerun.
Computer resources for reprocessing.
Reprinting costs.
Printer resource.
Additional paper costs.
Distribution resources to handle output.
Distribution and/or courier costs to deliver output.
Business degradation due to user wait.

Suggestion

With Control-D, the end users will be able to control their own output. The implementation of online viewing (described in Phase 9: Online Viewing Implementation) or Control-D/WebAccess Server (described in Phase 10: Control-D/WebAccess Server Implementation), would eliminate the resource drain described above. The user could simply view the report online, or if required, simply reprint the two unusable pages to the remote printer.

Information Required

Table 4 Current Level of Report Reruns

Item	Estimate
Number of report rerun requests raised:
Number of calls to help desk (re. reports):
Man hours each week for reruns:

What are the Average Report Delivery Times?

Many sites have now introduced Service Level Agreements (SLA) for report delivery. If you have such agreements, you will be aware of any exceptions to the specified report delivery times at your site. The objective for most sites is to have report output delivered to the user by the start of the working day. On some occasions, if processing problems occur or if printing volume is high, targets and SLAs will be missed. Report delivery may also be delayed if the user is located a considerable distance from the distribution center.

It is difficult to calculate what impact delayed reports have on the business. It depends on the report contents, and the need the user has for the report.

Let’s take the example of a bank. There is a particularly heavy printing load in the distribution center at end of month. Many users will not receive their output at the normal delivery times. One of the reports that has missed its dispatch deadline is the Lost/Stolen Check Book report. The bank manager for a certain branch has unwittingly cashed a stolen check because the updated report for that day was not received on time. If the report had been dispatched on time, the fraud could have been averted.

Suggestion

The implementation of online viewing (described in Phase 9: Online Viewing Implementation) or Control-D/WebAccess Server (described in Phase 10: Control-D/WebAccess Server Implementation) would allow the users access to their reports as soon as they are created, enabling them to meet their business deadlines.

Information Required

Table 5 Average Report Delivery Times

Item	Estimate
How often are report deadlines missed (%)?
What percentage of reports are delivered by 0830?
Number of inquiries about undelivered reports each week:

How Many Report Recipients Have Access to a Terminal or PC?

Many, but not all, users have access to some type of online facility. If you want to convert to online viewing under Control-D, you must find out if your current report recipients have access to online facilities or a PC.

Users in a remote branch use PCs extensively. You currently dispatch to them about three boxes of output per day. They do not want to use any of the available mainframe online applications to view report output, but they would like to view their reports on their PCs.

Suggestion

Using Control-D/WebAccess Server (described in Phase 10: Control-D/WebAccess Server Implementation) you can automatically download reports to a PC or file server so that the information is available for the users when they arrive at work.

Users without PCs can access reports under their preferred environment. The following environments are supported:

TSO
TSO/ISPF
ROSCOE/ETSO
CICS
VTAM
IMS/DC
IDMS/DC
COM-PLETE
VM/CMS PC

Information Required

Table 6 Recipient Access of Reports

Item	Estimate
Number of report recipients without online access:
Number of recipients with PC only:
Online environments used (CICS/TSO, and so on.):
List of reports currently viewed online:

What Happens to the Report Once Received by the User?

Often, once a report has been dispatched, we have no idea what the user does with it. It may be that the report is forwarded from a distribution point at the user’s location. Several users may reference the same report. Sections of the report may be split off and distributed within the user’s environment. The report may be used immediately, or it can be shelved for reference. The users may look for specific exceptions in a report or they may copy report details into a PC application.

A report is dispatched from the distribution center and delivered to a user department. The same report is accessed by five users, each looking for different exception values. One of the users then retypes figures from the report into a PC Spreadsheet application.

Suggestion

Using Control-D online viewing services (described in Phase 9: Online Viewing Implementation), users can access the same report concurrently and easily apply their individual exception criteria, eliminating a time-consuming and error-prone manual process. If reports are downloaded to their PCs using Control-D/WebAccess Server (described in Phase 10: Control-D/WebAccess Server Implementation), users can automatically copy report details into another PC application (such as spreadsheet applications).

You will need to know the user’s procedures to ensure that you deliver exactly their requirements using Control-D. These procedures will be revealed when you begin the user consultation process (described in Phase 8: Production Implementation).

You should by now have completed the distribution system review.

Decide Project Objectives

Now that you have completed the review of your current distribution system, you will start to make decisions about your implementation. To begin with, you must decide your project objectives. When deciding your project objectives, you should be realistic about your time frames and resources. For example, it is not realistic to target all suggested objectives, assign only one person for two hours per day, and then expect that the project will be completed within two weeks. All these factors must be taken into account when planning your implementation.

Use the following list to indicate your objectives for the project. After each objective, the corresponding phase of the implementation plan is listed.

Define recipient tree (Phase 2: Define Recipient Tree).
Design decollating missions (Phase 3: Design Decollating Missions).
Collate user reports into one printed bundle (Phase 4: Design Print Bundles).
Produce user report bundles on remote printers (Phase 4: Design Print Bundles).
Archive report output (Phase 5: Implementation of System Administration Tasks).
Automatically control printer workload balancing (Phase 5: Implementation of System Administration Tasks).
Eliminate report reruns (Phase 5: Implementation of System Administration Tasks).
Reduce size of spool volumes (Phase 6: Implement CDAM Direct Write).
Handle MSGCLASS output (Phase 7: Handling MSGCLASS Output).
Reduce volume of printed output (Phase 8: Production Implementation).
Eliminate redundant reports (Phase 8: Production Implementation).
Deliver online viewing services to end users (Phase 9: Online Viewing Implementation).
Implement report security (Phase 6: Implement CDAM Direct Write and Phase 9: Online Viewing Implementation).
Download packaged reports to the PC environment (Phase 10: Control-D/WebAccess Server Implementation).
Review project (Phase 11: Project Review).

You can target any or all of the above objectives. To achieve all of the above objectives, follow the supplied implementation plan in its entirety.

You can omit from your implementation plan any objectives you want. However, achievement of objectives requires implementation of the corresponding phases. When deciding which objectives to implement or omit, be aware that we strongly recommend that certain phases be implemented by all users, while we consider other phases to be optional.

Phases 1–3, 5−6, 8, and 11 are strongly recommended for all users. You should be able to achieve most or all of their corresponding objectives as part of initial implementation.

The following are the optional phases of the supplied plan:

Table 7 Optional Phases

Phase	Objective	Detail
Phase 4	Design Print Bundles	Can be skipped if implementing online viewing only.
Phase 7	Handling MSGCLASS	Can be skipped if you do not want to handle JCL output.
Phase 9	Online Viewing Implementation	Can be skipped if you only want to bundle and print user output.
Phase 10	Control-D/WebAccess Server Implementation	Can be skipped if you do not want to download reports to a PC.

Targeted objectives do not necessarily have to be achieved during initial implementation. You may be limited by time constraints or resources. If so, omit any of the optional phases from your initial implementation and tackle them later.

You may also have additional objectives that you want to add to the plan, and you should feel free to do so.

Recommendation

BMC recommend that you implement the plan in its entirety. If you are restricted by time constraints and resources, then we recommend that the only phases you skip for the initial implementation are Phases 7 and 10. These are the optional phases indicated on the project flowchart diagram (Figure 1). Although Phases 4 and 9 are considered optional, BMC strongly recommend that you do not skip them unless you are absolutely sure of the implications of doing so.

You should by now have decided your project objectives.

Choose a Pilot Application

Now that you have determined project objectives and selected the phases that you will implement, it is time to plan your strategy. Begin by trying to select an application that would benefit most from automation.

Use the following list to identify the most suitable application for the pilot implementation:

An application that has around 40 batch jobs.
An application that produces around 100 specific reports (preferably with reports for multiple users).
An application department with around 30 report recipients.
An application with non-critical reports.
An application whose users are dissatisfied with current distribution service.
An application whose users are asking for online viewing capabilities.
An application whose users are asking for program amendments to report formats.
An application with moderately-sized reports.
An application with uncomplicated decollating and bundling requirements.
An application with delivery locations that are far from the distribution center and would benefit from online viewing facilities.
An application whose users receive many reports from different jobs within the application.
An application whose users require a checklist of distributed reports.
An application department interested in function automation.
An application department with online access to mainframe systems.
An application department with remote printer facility.
An application department aiming for a "paperless office" approach.

Recommendation

We recommend that you select a pilot application that best fits the above criteria, and take it through all the recommended phases of the implementation plan. Of course you can follow any strategy you choose. It may be that you prefer to implement bundling and printing across all applications before you deliver any online viewing services.

You should by now have selected your pilot application.

Select Super Users

You have selected your pilot application. Now you should select your super users.

What is a Super User?

We use the term super users as a metaphor for the end users who will be your link to the application departments. They are a strategic resource for the project. The implementation of Control-D ultimately affects the report recipients. The super users will enable you to get feedback during the implementation and will drive the project in the end user environment.

Super User Characteristics

A super user should be in a position of authority in the end user environment.

Super users must be sold on the benefits that Control-D will provide to their organization. They must feel enthusiastic that their department has been selected for the pilot implementation, and that they will be the first users to receive the benefits of such an implementation. Their enthusiasm is critical in gaining the cooperation from the end users in the department. If you cannot sell the benefits of Control-D to the super user, then you have little chance of success in reaching your implementation objectives.

The Super User’s Role

The super users will sell the benefits of the implementation downwards through their organization. A super user who is sufficiently high up in the organization will probably deputize staff whom you can contact and with whom you can coordinate the project. These, in effect, will become your super users. Having super users appointed by their managers will provide a good basis for the recognition of the project and its subsequent implementation. Throughout the project you will be dealing with these users; they are strategic to your implementation and should ensure that your requests are met.

Suggestion

You might begin by approaching the application support team for information regarding the department organization, and for the names of potential super users. To get high-level management commitment, you may need your own managers to assist you in selling to the managers of the target application.

You should by now have selected your super users.

Assign Resources

Once you have decided upon your implementation strategy, you should then assign resources for the project. We will be looking at two main areas of resources:

Human resources.
System resources.

What Human Resources Do I Need?

Having identified your objectives for the project, and considered any time frames that you must meet, examine what human resources are required for the implementation. It would be ideal if you had unlimited human resources available to assign to the project, but in reality this seldom occurs. It may be that you are solely responsible for the implementation, or that someone has been assigned as a resource without regard to your objectives. If so, this will obviously impact your time frames.

Different people have different qualities and experience. If you have a junior member of staff assigned, it may take that person twice as long to do certain tasks as a more experienced person. You should factor this into your estimations.

What Skills are Required?

Certain skills are required for certain phases. In the list that follows, you are the project management resource and, as you can see, you are involved in all phases. The operations resource will be someone with knowledge of JCL, MVS, JES, and so on. This may be a production control or operations support person. The distribution resource may be any dispatch staff that you think may be able to learn and use new technologies (if this is not applicable then you can assign more operations resources). The end user resource will normally be your super users. The list below shows what skills may be involved in each phase:

Table 8 Skill Requirements

Action	Human Resources Involved
1. Deciding implementation strategy	Project Management
2. Defining the recipient tree	Project Management Operations Distribution End User
3. Designing the decollating missions	Project Management Operations Distribution
4. Designing print bundles	Project Management Operations Distribution
5. Implementing system administration tasks	Project Management Operations
6. Implementing CDAM Direct Write facility	Project Management Operations
7. Handling MSGCLASS output	Project Management Operations
8. Production implementation	Project Management Operations End User
9. Online Viewing implementation	Project Management Operations End User
10. Control-D/WebAccess Server implementation	Project Management PC Support End User
11. Project review	Project Management

Recommendation

BMC recommend that you involve as many staff members as possible in the project. This will help expand the knowledge base of Control-D at your site, and will mean that the project is not dependent on one or two individuals, which minimizes the impact if your Control-D resource leaves, or is sick or on holiday.

If you have decided to use the distribution staff as part of the implementation team, ensure that they have the motivation and ability to learn and use new technologies. If you can provide some basic training and set down standards, there is no reason why the distribution staff should not be a productive asset for the implementation. It will also provide you with an opportunity to introduce the distribution section to Control-D, and provide them with a feeling of involvement in the automation of their current functions.

The quality of the implementation team, and its motivation, can make the difference between success and failure of the project.

How Long will the Project Take?

If you selected a pilot application conforming to the suggested sizes, we recommend the following estimated time frames for each phase of the implementation plan. (Time frames are assigned in working days.)

Table 9 Estimated Time Frame For Each Phase

Action	Man Days Required	% of Project
1. Deciding implementation strategy	2 Days	11%
2. Defining the recipient tree	1 Day	6%
3. Designing the Decollating Missions	4 Days	21%
4. Designing print bundles	1 Day	6%
5. Implementing system administration tasks	1 Day	6%
6. Implementing CDAM Direct Write facility	1 Day	6%
7. Handling MSGCLASS output	1/2 Day	3%
8. Production implementation	1 & 1/2 Days	9%
9. Online Viewing implementation	3 Days	15%
10. Control-D/WebAccess Server implementation	2 Days	11%
11. Project review	1 Day	6%
TOTAL	18 Days	100%

These are estimates of the actual amount of man days required. It is not a total for the actual elapsed time of the project as there are periods when the users will be adjusting to their latest changes before we provide them with another benefit. It also does not take into account weekends, holidays and any unscheduled interruptions to normal activities.

You can use the percentage breakdowns if you want to vary the scope of the project according to size of application or experience level of staff involved.

Do not use the initial implementation time frames as a forecast for the entire implementation of Control-D. These time frames apply only to the initial pilot application. It is important to understand that any subsequent applications will be implemented in a fraction of the time. The initial implementation is an educating phase where you will be setting foundations for all other applications. Also, some phases will not have to be carried out for future applications, further reducing the implementation time.

The time frames supplied are estimates. You may find that the pilot implementation takes more or less time, depending on the specific requirements and complexity of your site.

Assign Project Tasks

You now have an idea of what skills are required, how long each phase will take, and what resources you have for the project. Now you should assign tasks to the relevant individuals. You will see from the project flowchart that some phases can be worked on at the same time. If you have enough resources, you can assign them to the various phases, and multi-task the project.

Assume that you have been assigned one operations person, and two people from the distribution staff. After Phase 2, "Designing the Recipient Tree," you could assign the operations person to "Handling MSGCLASS Output" (Phase 7) while the distribution staff set up the Decollating Missions (Phase 3).

Recommendation

BMC recommend that you be aware of what is involved in each phase of the project so that you can direct operations. It is your responsibility to set up standards for implementation, and to ensure that these standards are followed. You will constantly review all work carried out on the project to ensure that each phase is being implemented according to your directions.

What System Resources Do I Need?

During the installation process, sizes are allocated for the Control-D repository files according to calculations provided in the INCONTROL for z/OS Installation Guide. These will probably be sufficient for your pilot application, but it is recommended that you review them. Another important parameter that is set during installation is the volumes to be used for CDAM files. All reports handled by Control-D will be stored in this compressed format. Ensure that there is enough space to handle all the reports of your pilot application. In Phase 5 we will discuss how we back these files up and how long they should stay on DASD.

You should by now have assigned resources for your implementation.

Review

During this phase you have studied your current distribution environment to see how it operates. You have also examined the potential project objectives and implementation strategies, and reviewed the resources and time frames for the project.

Before you continue, you should have:

Performed a distribution system review.
Decided your project objectives.
Selected your pilot application.
Selected your super users.
Assigned resources to the project.