|
Problem IC76144
|
Status: Closed |
ASSERTION WHEN REATTACHING PARTITIONS AFTER DROPPING LOCAL INDEX IN A DETACHED PARTITION |
| product: |
DB2 FOR LUW / DB2FORLUW / 970 - DB2 |
| Problem description: |
Dropping local indexes is not correctly handled by the index
partition on a 'D' (detached) partition or a 'L' (logically
detached) partition. Local indexes expected to be dropped on
these partitions could be left in the index object. Most of the
time, this not a problem as the result tables are usually
deleted. However, if the partition is re-attached back to the
original table at a future time, you may see an assertion during
ATTACH.
This problem only occurs if the drop local index happens on
table that has partitions in 'D' state (i.e. there are deatched
dependent tables to be maintained) or in 'L' state. This
scenario should be rare as drop index is typically a planned
database design operation, it is unusual for one to drop indexes
while performing data maintenance on the table.
When this assertion occurs while attempting to re-attach the
partition, the following error will appear in the db2diag.log:
2010-08-06-09.42.36.860067-240 I6781546E5616 LEVEL:
Severe
PID : 16614 TID : 47199689894208 KTID :
17605
PROC : db2sysc 0
INSTANCE: spilchen NODE : 000 DB : TEST1
APPHDL : 0-53 APPID: *N0.spilchen.100806133552
AUTHID : SPILCHEN
EDUID : 54 EDUNAME: db2agent (TEST1) 0
FUNCTION: DB2 UDB, index manager, sqliUpdInxTypForAttach,
probe:50
MESSAGE : ZRC=0x87090054=-2029453228=SQLI_PRG_ERR "Program
error"
DIA8575C An index manager programming error occurred. |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* All *
****************************************************************
* PROBLEM DESCRIPTION: *
* A small timing hole caused the index to not be dropped. *
****************************************************************
* RECOMMENDATION: *
* Upgrade to DB2 V9.7 Fix Pack 5. *
**************************************************************** |
| Local Fix: |
Do not drop local indexes while simultaneously detaching
partitions or while there are detach dependent MQTs that need to
be maintaned. Check the partition status from
SYSCAT.DATAPARTITIONS to make sure that no partitions in 'D' or
'L' states before dropping local indexes from the table.
If the drop local index has already happened, the user can query
SYSCAT.INDEXPARTITIONS to find out the tablespace ID and the
object ID of the specific index partition and mark it invalid
using DB2DART with the /mi option. Afterwards, reissue the
ATTACH command. The local index will be correctly rebuilt by DB2
during SET INTEGRITY after attach. |
| available fix packs: |
DB2 Version 9.7 Fix Pack 5 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 6 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 7 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 9a for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 10 for Linux, UNIX, and Windows
|
| Solution |
This problem was first fixed in DB2 V9.7 Fix Pack 5. |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 03.05.2011
07.12.2011
07.12.2011 |
| Problem solved at the following versions (IBM BugInfos) |
9.7.FP5
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 9.7.0.5
|
 |
