The following points need to be true to run into this situation
during rollforward:
* Backup needs to take a long time to complete for a tablespace
* An additional space map page needs to be allocated to the
index object during backup before truncate table
* Index object needs to change (truncate table) after object
table has been archived but before index object is archived
* Index object needs to contain more than one space map page
The rollforward of the additional space map page after the
restore, before the truncate table, will fail with the following
messages in the db2diag.log file:
2014-02-15-06.59.39.779236+000 I2200215A525 LEVEL: Severe
PID : 21626960 TID : 13462 PROC : db2sysc
0
INSTANCE: xxxxx NODE : 000 DB : yyy
APPHDL : 0-3427 APPID: *LOCAL.xxxxx.140215062931
AUTHID : xxxxx
EDUID : 13462 EDUNAME: db2redow (yyy) 0
FUNCTION: DB2 UDB, index manager, sqliRedoInitSMP, probe:2693
RETCODE : ZRC=0x87090054=-2029453228=SQLI_PRG_ERR "Program
error"
DIA8575C An index manager programming error occurred.
...
2014-02-15-06.59.39.824851+000 I2201631A525 LEVEL: Severe
PID : 21626960 TID : 13462 PROC : db2sysc
0
INSTANCE: xxxxx NODE : 000 DB : yyy
APPHDL : 0-3427 APPID: *LOCAL.xxxxx.140215062931
AUTHID : xxxxx
EDUID : 13462 EDUNAME: db2redow (yyy) 0
FUNCTION: DB2 UDB, index manager, sqliRedoInitSMP, probe:4950
RETCODE : ZRC=0x87090054=-2029453228=SQLI_PRG_ERR "Program
error"
DIA8575C An index manager programming error occurred. |
