Managing IMS Data Entry Databases

Analyzing DEDB areas

As root segments and dependent segments for database records are added, updated, or deleted in a DEDB area, those segments can become scattered across a large number of separate CIs, or the free space in CIs can become fragmented into smaller pieces. The scattered segments and fragmentation of free space might cause the performance of applications to suffer and lead to inefficient use of space. Also, when available free space is used up, new segments cannot be added.

Reorganizing or restructuring DEDB areas is required to prevent degradation of application performance and out-of-space conditions.

IMS Fast Path Solution Pack provides both online and offline area analysis functions through IMS Fast Path Online Tools and IMS Fast Path Advanced Tool.

The online analysis function that is provided by the Online Pointer Checker (OPC) utility of IMS Fast Path Online Tools and the offline Analyze function of IMS Fast Path Advanced Tool verify the integrity of pointer values and analyze the space utilization, I/O efficiency of root segments and database records, and randomizing efficiency of a DEDB. Both online and offline analysis functions can process multiple areas of a DEDB in a single job step. These analysis functions can also reveal potential pointer errors.

These analysis functions generate area analysis reports that contain various statistics about the areas. The reports can help you reduce the amount of time that you spend diagnosing and solving DEDB space management issues and problems.

You can use the online analysis function if you want to analyze DEDB areas without taking the DEDB or areas offline. The online analysis function uses the buffering and locking services of IMS Fast Path system to read segments from online DEDB areas while maintaining integrity.

You can use the offline analysis function if you want to process multiple DEDB areas with high performance. The offline analysis function can process multiple areas in parallel in the job step by using its own high-performance I/O technology. The offline analysis function can also use image copy data sets as input instead of area data sets. By using image copy data sets as input, you can analyze DEDB areas without affecting the online applications service schedule.

Managing space of online areas

Taking an area offline only for the purpose of reorganizing free space is often not a viable option in DEDBs, especially in data centers where IMS environments are kept online 24 hours a day and 7 days a week. IMS provides the High-Speed DEDB Direct Reorganization utility (DBFUHDR0) to reorganize free space of online DEDB areas.

The Online Expert Reorganization (OER) utility of IMS Fast Path Online Tools extends the capability of DBFUHDR0. OER can evaluate UOWs in DEDB areas, rank the UOWs, and reorganize only highly ranked UOWs to reduce the volume of database update logs. OER can also collect statistics and generate area analysis reports of the reorganized area by calling the online area analysis function (OPC) after the reorganization process. The statistics data and analysis reports provide information about how effectively the free space was reorganized and whether any space utilization or I/O efficiency issues remain in the reorganized area.

Preventing an out-of-space condition is also crucial in operating environments that cannot be stopped. If space in the independent overflow (IOVF) portion or the sequential dependent (SDEP) portion is about to be used up, and if the database service cannot be stopped, the Online Area Extender (OAE) utility of IMS Fast Path Online Tools can help you dynamically extend the size of the DEDB areas. You can specify the portion of the area to extend: the IOVF portion, the SDEP portion, or both.

These utilities can be run by as stand-alone utilities, but they can also be run by using the Online Space Management utility. The Online Space Management utility integrates the online reorganization process of OER, the online area extension process of OAE, and the online area analysis process of OPC into one job step. When the Online Space Management utility is used in the active autonomics environment, it runs one or more of these processes based on the state of the area; whether the area requires reorganization, area extension, or both. This utility simplifies space management tasks and eliminates manual operations that are needed to individually run these utilities.

Tuning DEDB areas and randomizer

As database records and segments in an area increase, the space utilization of the area becomes inefficient to the point that tuning is required. A poorly tuned DEDB area might require excessive physical I/Os to read root segments and database records.

However, tuning a database often involves significant and time-consuming database administration tasks including tuning the randomizer, increasing the amount of free space, changing the database structure, or adding areas.

The Tune function of IMS Fast Path Advanced Tool can help ease these tuning tasks. The Tune function facilitates multiple iterations of the database tuning process, which allows you to select physical database attributes that meet your performance and space utilization requirements.

This function helps you determine the effective changes to database definition and randomizer by evaluating the potential benefits, or impact, of the changes to the area without the overhead of unloading and reloading the area. By ensuring that your new database definition is effective before reorganizing your database, you can avoid rerunning reorganization jobs that might be caused by inadequate database definitions.

Restructuring offline areas with high performance methods

When you have tuned the database definition or the randomizer, the subsequent steps are unloading the segment data from the existing area, reclaiming the space of the area based on the new database definition parameters, and reloading the unloaded data to the reclaimed space, possibly with the new randomizer. These tasks typically require you to take the database or the area offline for a short period of time.

By using the Unload, Reload, Change, and Reorg functions of IMS Fast Path Advanced Tool, you can perform the offline reorganization processing efficiently.

Unloading and reloading areas in multiple steps

Both the Unload and the Reload functions use their own high performance I/O technique to reduce CPU time and elapsed time. If you combine the unload step and the reload step, a DEDB area can be read and loaded faster than when the area is processed by IMS online applications.

This method creates an intermediate file (a sequential file that contains unloaded segment records) that can be used to obtain specific database statistics for data analysis or that can be used to create DEDB areas for testing purposes.

Changing an area in a single step

The Change function does not create an intermediate file. The unloaded segment data is passed to the reload process through the main storage buffers. The internal unload and reload processes use the same high performance I/O technique that is used by the Unload and Reload functions. Through the use of these techniques, the Change function further reduces the elapsed time of the entire reorganization process.

Integrating image copy process

The Unload, Reload, Change, and Reorg functions of IMS Fast Path Advanced Tool can internally call an image copy process. The Unload function supports unloading of segment data from image copies. This capability can reduce impact on online database applications because it does not read the area data sets again for batch application data processing or reorganization processing. You can take image copies of reloaded areas by using IMS High Performance Image Copy for z/OS® within the Reload, Change, or Reorg function job step.

Restructuring online areas

If you use the Online Structure Change utility, you can restructure online DEDB or areas with minimal impact to database availability. With this utility, you can change DBD descriptions, randomizing modules, and segment edit/compression routines.

Taking systems offline for maintenance can affect your business. In today's market, systems must be highly available because the cost of outage can be too high. Online restructuring provides users with full access to the database during database restructure, improves the overall database availability, and reduces planned downtime.

The utility allocates shadow area data sets to perform unload and reload while capturing all database updates made by applications. When unload and reload is done, it uses an IMS DBCTL system to apply all the captured updates and switches the original area data sets with new restructured area data sets. A single outage is necessary to switch the area data sets. Also, when PRESTOP=YES is specified, an additional outage occurs during the initialization phase.