Evangelizing Mainframe
Print Email

IMS 13 Aims to Reach New Heights

While all the excitement might be around big data and Hadoop, the reality, for most mainframers, is that their largest amounts of data are still being stored in IMS hierarchical databases being accessed quickly and accurately all day long. For those organizations, enhancements to IMS can be very important. The Virtual IMS user group recently had a presentation titled, “Taking IMS to New Heights—What the Future Holds for IMS,” from IBM IMS Chief Architect Betty Patterson. Here’s just some of what she had to say:

Patterson started her presentation by introducing IMS V13, and saying it was smarter than ever. IMS 13 has the highest efficiency and the lowest total cost of ownership.

In terms of reducing costs, the IMS 13 developers had a cross-platform focus on reducing mainframe software costs, a major focus on reducing CPU usage, and they made changes throughout IMS to improve performance, reduce storage usage, and reduce CPU used to run IMS. IMS 13 offers:

• More efficient storage services
• Improved algorithms
• Reduced path length
• Optimized frequently used processes
• Latch/lock improvements
• Storage reductions, and
• Use of System z hardware functions

The benefits include improved performance, lower cost per transaction, and reduced cost of ownership.

Patterson was soon looking in detail at the nuts-and-bolts level, explaining how the new Open Transaction Manager Access (OTMA) early termination notification allows it to leave the XCF group earlier in the termination processing. It notifies OTMA clients (IMS Connect, IBM WebSphere MQ, OTMA Callable Interface) of the termination via their XCF Group Exit. Clients can then route requests to other systems, which addresses issues associated with transaction messages being accepted but not processed. The benefits include autonomic enhancement for higher availability that allows OTMA clients to be informed of an IMS shutdown in order to choose more timely alternatives. It potentially reduces unsuccessful attempts to send in new requests and can expedite shutdown processing.

The system definition macro SECURITY is no longer used as part of the IMS system-generation process. You specify security settings through PROCLIB members’ RCLASS parameter added to DFSPBxxx. RCLASS is also supported in DFSDCxxx. The DFSPBxxx RCLASS parameter value overrides DFSDCxxx if both are specified. The SECCNT parameter has been added to DFSDCxxx. Other security settings continue to be specified in DFSPBxxx, including SECLVL (parameter is replaced by RCF, TRN and SGN in DFSPBxxx) and TYPE (parameter is replaced by ISIS in DFSPBxxx). The benefits are a simplified system generation process, and an easier method to update security-related settings.

In terms of IMS 13 integration, a new option supports TCP/IP network connectivity for intersystem communication (ISC) connections. The benefits are that it provides a strategic protocol alternative to SNA/VTAM, allowing an all-inclusive TCP/IP solution for networks.
Synchronous program switch extends IMS Synchronous Callout to invoke another IMS application. Synchronous flows use IMS Data Language/I (DL/I) ICAL, asynchronous flows still use DL/I ISRT. The OTMA Descriptor defines the destination. This provides a single DL/I call to request a synchronous service regardless of where that service resides, and it simplifies integration and improves usability.

Asynchronous Callout to WebSphere MQ via MQ Bridge uses OTMA Descriptor enhancements. The new TYPE=MQSERIES defines a WebSphere MQ destination. It provides asynchronous callout and messaging support (DL/I ISRT ALTPCB)—a new option that allows exits to be called to override the descriptor. It applies to all destination descriptors. This eliminates the need to write an OTMA user exit to recognize an MQ destination, and simplifies integration and improves usability.

Patterson informed the user group that IMS 13 allows Java-dependent regions to use the External Subsystem Attach Facility (ESAF). It allows connections for DB2 to be consistent across all region types. It allows access to other subsystems such as WebSphere MQ and eliminates the need to use z/OS Resource Recovery Services (RRS) attach for DB2. The benefits are that it allows Java Messaging Service (JMS) access to MQ from Java. It allows MQ access from COBOL and PL/I. It simplifies external subsystem definitions. And there’s improved performance for DB2 due to eliminating extra sign-on processing.

Among the IMS 13 core capabilities is High Availability Large Database (HALDB) Alter. This allows users to change the structure of an IMS HALDB without a database outage. Users can add a new field to space at the end of an existing segment, increase the length of an existing segment, and define new fields that redefine (overlay) existing fields and space in an existing segment. It improves IMS HALDB availability by providing structure changes without taking the database offline, and provides flexibility in rolling database changes into the system.

The Fast Path Data Entry Database (DEDB) Alter allows DEDB area changes without an unload/reload of the area. It allows dynamic change for UOW and ROOT parameters of an existing area. It replaces the randomizer. It provides a new DEDB dynamic change utility that runs as a standard Fast Path utility, and the area remains online. The benefits are that it improves DEDB area availability by providing definitional changes without taking the area offline, and it provides flexibility in rolling area changes into the system.

Database versioning allows programs to use different versions of the same physical database, providing multiple views of the physical data maintained in the IMS catalog. Existing applications can remain unchanged when the physical structure of the database changes. Users can recompile just those programs referencing changed fields/segment. This applies to full-function databases, HALDB and Fast Path DEDB. It supports HDAM, HIDAM, PHDAM, PHIDAM and DEDB databases. It provides greater flexibility in rolling out new versions of programs and databases, and allows new programs to get out faster without waiting for all programs to be updated to the new database structure.

The limit of concurrent application threads has been increased to 4095. The limit applies to the total number of combined dependent regions, CICS/DBCTL threads and open database access (ODBA) threads. It follows a change to the MAXPST parameter on the IMS control region. It offers increased capacity and scalability for IMS systems, allows vertical growth, and provides more dependent regions for use with synchronous callout and program switch.

Refreshable user exits mean users can refresh user exits without an IMS system outage, and define exit “types” to support a list of exit names. This improves availability and simplifies user exit management. IMS Connect enhancements mean users can dynamically CREATE IMS Connect resources through commands for PORT and DATASTORE.

In conclusion, Patterson said the IMS strategic objectives were to reduce total cost of ownership, reduce MIPS usage and use advanced autonomics, so IMS is self-managing and self-tuning.
IBM wants to extend the IMS lead in availability, scalability and performance. And IBM wants to consistently deliver IMS capacity limits that are well beyond customer needs and performance metrics that help users grow their business securely. It seeks to extend the lifecycle of IMS applications and transactions, and enable high-volume transaction processing for the next wave of applications (big data, next generation Web services, cloud, mobile, and more).

Trevor Eddolls is CEO at iTech-Ed Ltd., an IT consultancy. For many years, he was the editorial director for Xephon’s Update publications and is now contributing editor to the Arcati Mainframe Yearbook. Eddolls has written three specialist IT books, and has had numerous technical articles published. He currently chairs the Virtual IMS and Virtual CICS user groups.

Posted: 7/23/2013 1:01:01 AM by Trevor Eddolls

Print Email

Join Now!