DB2 FOR LUW / DB2FORLUW / 970 - DB2

An instance may abend when copying data into a table queue under 
some very specific access plan conditions. 
A typical trap file may show the following functions on top in 
the stacktrace : 
 
Stack: 
====================== 
 
 memmove 
 sqlkt_pack_tuple 
 sqlktins 
 sqlritqb2 
 
 
 The problem is due to an incorrect buffer location being 
accessed. 
 This can only be confirmed by db2 service personnel by 
examining the "section dump" of the execution plan, 
 which is located in the binary dump file. 
 
 From an access plan point of view, at least the following 
conditions need to be true  : 
 * There is a star join that has a join of a dimension table 
(above the IXAND operator ) that references 
 the same columns of the dimension table that were already 
accessed in the star join (below the IXAND operator ) 
 ( this may also be referred to as a back join ) 
* The dimension table column that is accessed in the star join 
is accessed via list prefetching 
  i.e. there is a RID scan operator followed by a fetch 
operation on the base table. 
 
* The index scan in the list prefetch of the dimension table 
within the star join has a sargable predicate on a column that 
is used in the back join.

**************************************************************** 
* USERS AFFECTED:                                              * 
* all users with a DPF environment                             * 
**************************************************************** 
* PROBLEM DESCRIPTION:                                         * 
* An instance may abend if this apar is not applied. Although  * 
* the conditions for triggering this problem are rare, there   * 
* is no easy way to predict if your workload is liable to this * 
* problem.                                                     * 
**************************************************************** 
* RECOMMENDATION:                                              * 
* Upgrade to v9.7 Fix Pack 1                                   * 
****************************************************************

There is no known workaround for this problem, other than to 
force a different access plan to avoid 
 at least one of the above conditions.

DB2 Version 9.7 Fix Pack 1 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 2 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 3 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 3a for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 4 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 5 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 7 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 9a for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 6 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 8 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 9 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 10 for Linux, UNIX, and Windows
 

Problem was first fixed in Version 9.7 Fix Pack 1

keiner bekannt / siehe Local-Fix

Datum - Problem gemeldet    :
Datum - Problem geschlossen :
Datum - der letzten Änderung:

18.06.2009
22.02.2010
22.02.2010

9.7.FP1