APAR status
Closed as program error.
Error description
APAR PM82937 provided a new function to specify a padding character when segment length is increased due to a DBD change. The padding process is invoked even when segment length is not increased and the process works incorrectly when a variable length segment or a compressed fixed length segment is defined. When a variable length segment or a compressed fixed length segment just fits the end of a block, the 1st byte of the FSEAP of the next block is broken by the padding character. (The default padding character is X'00'.)
Local fix
Problem summary
**************************************************************** * USERS AFFECTED: All users of IBM IMS High Performance Load * * for z/OS Version 2 Release 1 (FMID=H1IM210). * **************************************************************** * PROBLEM DESCRIPTION: Applying new function of APAR PM82937 * * can break an FSEAP. * **************************************************************** * RECOMMENDATION: Apply the maintenance for this APAR. * **************************************************************** This APAR fixes the following two issues raised by APAR PM82937: 1. Applying the APAR PM82937 can break an FSEAP, and a data loss can be caused by the broken FSEAP. 2. WF1 records contain garbage when padding charecter other than X'00' is specified in the PAD statement. The details are as follows: 1. APAR PM82937 provided a new function to specify a padding character when segment length is increased due to a DBD change. The padding process is even invoked when segment length is not increased and one of the following conditions is satisfied: - A segment edit/compression exit is used and the length of fixed-length segments is shorter than the maximum number of bytes that is defined by the COMPRTN parameter in the DBD definition. - The length of variable-length segments is shorter than the minimum segment length that is defined in the DBD definition. When the padding process is invoked and any of these segments are loaded right at the end of a block, the 1st byte of the FSEAP of the next block is broken by the padding character. (The default padding character is X'00'.) Inserting segment by using an incorrect FSEAP can cause loss of data. 2. WF1 records contain garbage characters when all the following conditions are met: - APAR PM82937 has been applied and padding character other than X'00' is specified in the PAD statement. - Logical relationship or secondary index is defined for a compressed fixed-length segment and its length is increased during reorganization. - Unloaded data set is created in decompressed format.
Problem conclusion
======================== = Documentation Change = ======================== IMS High Performance Load for z/OS Version 2 Release 1 User's Guide SC18-9222-05 ( SC18922205 ) should be changed as follows: ================================================================ 2.1.5 Control statements : 2.1.5.x PAD control statement | | Specifies the padding character to fill out the segments to | adjust to the length defined in the DBD definition. | | The segments are padded in the following cases: | When a segment edit/compression exit is not used and the | length of fixed-length segments is increased due to a change | in the DBD definition. | | When a segment edit/compression exit is used and the length | of fixed-length segments is shorter than the maximum number | of bytes that is defined by the COMPRTN parameter in the DBD | definition. | | When the length of variable-length segments is shorter than | the minimum segment length that is defined in the DBD | definition. | This control statement is optional. If you do not explicitly specify this control statement, PAD=X'00' is used. _________________________________________________________ | | | | | >>__PAD=X'nn'____________________________________>< | | | |_________________________________________________________| nn | Specify a two digit hexadecimal code as the padding | character. For example, PAD=X'FF'. | | When all of the following conditions are met, the difference | in the segment length is padded with X'00' regardless of the | value specified on the PAD statement: | Segments are defined as fixed-length segments. | The use of a segment edit/compression exit is defined for | the segments. | The length of the segments is increased due to a change in | the DBD definition. | The unloaded data set is created in decompressed format. |
Temporary fix
********* * HIPER * *********
Comments
APAR Information
APAR number
PI05052
Reported component name
IMS HIGH PERF L
Reported component ID
5655E0700
Reported release
210
Status
CLOSED PER
PE
YesPE
HIPER
YesHIPER
Special Attention
NoSpecatt
Submitted date
2013-10-29
Closed date
2013-11-14
Last modified date
2013-11-14
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
UI12507
Modules/Macros
HPSRHDLA HPSRHDLB HPSRHDL2 HPSRHDL3 HPSRHDL7 HPSRHDL8 HPSRHDL9 HPSRHSLA HPSRHSLB HPSRHSL2 HPSRHSL3 HPSRHSL7 HPSRHSL8 HPSRHSL9 HPSRRLDA HPSRRLDB HPSRRLD2 HPSRRLD3 HPSRRLD7 HPSRRLD8 HPSRRLD9
SC18922205 |
Fix information
Fixed component name
IMS HIGH PERF L
Fixed component ID
5655E0700
Applicable component levels
[{"Business Unit":{"code":"BU048","label":"IBM Software"},"Product":{"code":"SSCX89F","label":"IMS HP Load"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"2.1.0","Edition":"","Line of Business":{"code":"","label":""}}]
Document Information
Modified date:
14 November 2013