What is IAM?
The IAM product is different from many other VSAM performance enhancement products. Rather than trying to tweak VSAM to work faster, the IAM approach starts off by using a different underlying file structure and data organization. This innovation, coupled with performance oriented algorithms and coding techniques, results in significant proven performance benefits.
The IAM product is being used to achieve the following benefits:
- Reduction of computing costs
- Reduction of CPU time
- Reduction of EXCPs
- Reduction of DASD space utilization
- Reduce nightly batch processing window
- Delay processor and DASD upgrades
The IAM product is available as a base product that supports KSDS, ESDS, and RRDS type of files. IAM provides support for Alternate Indexes (AIX) as well. The IAM/PLEX feature provides Sysplex record level sharing support. A free trial version of IAM is available for users to evaluate the IAM product. If requested it will also include the IAM/PLEX optional feature.
IAM Base Product
The base IAM product provides a unique set of features and functions to enhance and improve the performance of VSAM KSDS, RRDS, and ESDS types of files. These key features and functions include:
- A transparent VSAM interface that requires no changes to application programs, and generally no changes to JCL or CICS regions
- IAM files can easily exceed 4 gigabytes. A single IAM file has a maximum size of 54 terabytes. There are no special size considerations for KSDS files. For ESDS files to exceed 4 gigabytes, they must be defined with either an 8-byte RBA (IAM XESDS) or with the PSEUDORBA feature of IAM due to the VSAM addressing architecture. The current actual size limitation is due to the z/OS DD (TIOT) restriction of a maximum of 59 volumes. For example, on 59 3390-54 volumes, the maximum is 3,186 (54g*59) gigabytes.
- IAM data sets are typically allocated as non-VSAM type of data sets, generally DSORG=PS.IAM data sets can also be defined as Large Sequential data sets (DSNTYPE=LARGE) or as DFSMS Extended Format data sets to utilize more than 64K tracks per volume. IAM files can reside on EAV volumes, either as DFSMS Extended Format data sets, as Large Format Sequential data sets, as Basic Sequential data sets, or as VSAM Linear data sets. IAM files that are DSORG=DA should be deleted and redefined to use DSORG=PS.
- IAM files defined with a Keylabel, imply that Pervasive Encryption use is desired and such files will be defined as VSAM Linear data sets. Such files must be defined with the SMS Extended Format attribute and the SMS Extended Addressability attribute.
- IAM's automatic Real Time Tuning offers dynamic buffer management techniques that are unsurpassed. With the Real Time Tuning TURBO mode, IAM responds faster to high I/O needs, increasing the number of buffers dynamically, eliminating the need for manual tuning for almost all files.
- A Data Compression facility that can be used for all types of IAM data sets, with less CPU overhead than is typical of VSAM data compression techniques. IAM supports both an efficient proprietary software data compression technique or hardware compression. Hardware compression can be used with either IAM generating a compression dictionary during file load, or with a customized hardware compression dictionary. Users may be able to achieve even greater data compression than can be achieved with standard VSAM hardware compression when they opt to build a customized dictionary as described in this space. Data compression also helps to reduce physical I/O required to process the data sets, because more data is contained in a physical block than when it is uncompressed.
- Record level sharing within a single z/OS image (LPAR), for concurrently executing batch jobs, CICS regions and TSO users
- Support for journaling and recovery that offers capabilities normally seen only with data base software
- Automatic space release after a file is initially loaded and dynamic secondary space adjustments
- Software data compressed files can be reorganized with FDRREORG or IDCAMS without decompressing and re-compressing the data. This reduces CPU time and I/O to perform the reorganizations, and it can reduce media usage when the sequential copy of the data is placed on DASD.
- A Data in Virtual feature called Dynamic Tabling, which keeps those records that are frequently required in a Data Space or in a virtual storage table within the users address space
- IAM’s run time report IAMINFO, which fully describes each IAM data set processed by a job step, along with a complete processing profile, including resource usage. This data can also be captured by SMF, with reporting available from programs provided with IAM, or customers can write their own SMF analysis and reporting programs to include the IAM provided data.
IAM/AIX
The IAM/AIX feature is included with the base IAM software product at no additional cost. IAM/AIX uses the full functionality and features described above on IAM base clusters with IAM alternate indexes and IAM paths.
IAM/PLEX
The IAM/PLEX additional cost feature provides record level sharing support of IAM data sets across multiple LPARs within a z/OS® Sysplex. IAM/PLEX builds upon IAM/RLS by utilizing the XCF Sysplex services and the Sysplex Logger to provide a record level sharing capability for IAM files within a Sysplex. IAM’s implementation is an evolutionary step forward from the basic single system IAM/RLS, that utilizes the IAM/RLS framework that has been in use for several years. The IAM/PLEX design is different from VSAM/RLS and is intended to enable users to take advantage of record level sharing without the complexity involved with setting up a VSAM/RLS environment. IAM/PLEX includes support for sharing between online (CICS/TS) regions and batch jobs.