LongEx Mainframe Quarterly - November 2014
In the 1980s, mainframes printed a lot of output. Forms, documents, reports, job output; you name it, it went on paper. Mainframe sites had large printers busily going through boxes and boxes of paper. Output management products have helped save many of these trees, providing a way of storing and viewing output online. Today there are many different products to manage output. So what are they, and what do they do?
Document vs Output
All output isn't necessarily equal. In general there are two main groups of output: Plain Text and Documents
On z/OS there is a lot of output from jobs and started tasks. Much of this output, particularly from production batch and system tasks, needs to be retained for IT staff. The z/OS syslog also has valuable information that IT staff will want to keep.
This output is quite easy to manage. It is plain text, and normally used only by IT staff. There are rarely compliance requirements for its retention, and clients do not need access. There can be a lot of this output, but it is rarely excessive.
Simple, plain-text reports can also be managed with these plain text products.
Reports and Documents
These are more complicated. They can have sophisticated formatting and presentation features like images, fonts and overlays. They can be produced as very large reports, with different users using (or permitted to use) different sections. Individual sections may need to be merged together as a single report for some individual users or groups.
Security for reports and documents may be more complicated, as some users may have access to only some portions of each report. Clients may need access to these documents, increasing the compliance and security requirements.
There will be more functionality needed. For example, browser-based viewing, automatic email notification, and indexing for content searching. As you can imagine, the amount of this output can be huge.
Software to store plain text output has been around for years, and there are many options. These include:
This software typically monitors the JES Spool, and automatically takes output depending on selection criteria like class or destination. This output will be automatically managed: saved, archived and deleted as needed. Security is included to restrict access to output, as are ISPF applications to browse and manage the output.
These products usually rely on systems features like DFSMShsm for report archiving to tape. However they usually include automatic deletion of reports when they are no longer needed.
Some products take a step towards document and report management with extra features. For example BMC Control-D provides report distribution and bundling features, and an additional Report Approval feature. It also provides web access via Control-D/WebAccess.
Document and Report Management
Software for z/OS document and report management has come a long way over the years. Today examples of mainframe-supporting document management software include:
These products include all the functionality of the plain-text output management products above, and add more. Let's take IBMs CMOD as an example.
CMOD can run on many different platforms. The z/OS CMOD monitors JES spool, and takes output depending on selection criteria. Output can also be sent directly to CMOD from batch. CMOD accepts PDF, plain text and AFP documents. CMOD uses z/OS OAM for automatic document archiving and retrieval. Documents are usually saved initially in DB2, and then moved disk and tape as they get older, depending on the configuration. Different reports can have different archival options.
Reports are viewed from a web browser, including AFP and PDF documents. Security is managed by the z/OS security product (e.g. RACF), with sufficient granularity to limit report access to groups or individuals. It includes search functionality by indexing documents as they are accepted. CMOD can also convert input data to formats like HTML, PDF and XML.
The optional distribution feature allows automatic document distribution via PDF file, web, email and print. It supports document bundling, splitting and merging: merging multiple documents or parts of documents for distribution.
CMOD can run on run on Windows, System i and a few UNIX variants. However it can also run exclusively on z/OS. However many of the competitors reduce their mainframe footprint and rely on non-mainframe components. For example ISIS Papyrus has a z/OS agent forwarding output to a non-mainframe server.
IBM CMOD provides all the functionality within the one product. Others may merge or require other products to offer similar functionality. For example, Macro 4 Columbus Z provides report acquisition, conversion and distribution features. It needs partner, non-mainframe Macro 4 Columbus products for report archival and online viewing. Similar CA-Spool provides SPOOL acquisition, printing, transformation and forwarding. However it relies on other CA products such as CA View for online viewing, and CA Deliver for report bundling and distribution.
Many sites have two output management products: one for job/started task and syslog output, and one for report management and distribution. Plain text products provide all you need for archival and online viewing of basic job output and plain text reports. Report management products build on this functionality for sophisticated report management. They add bundling and distribution features, AFP document support, indexing, web based access, document translation and automatic notification.