DB2 FOR LUW / DB2FORLUW / B10 - DB2

There is a boundary condition during online backup where the
database on some member is in the process of performing log file
registration.
If the database encounters this timing window, the online backup
fails with error (SQL0901N) and you
may see db2diag.log messages like bellow.

----------------------------------------------------------------
2018-12-21-00.02.46.825839+540 I3632A1604 LEVEL: Severe
PID : 37093746 TID : 65329 PROC : db2sysc 1
INSTANCE: xxxxxxx NODE : 001 DB : XXXXXXXX
APPHDL : 3-276 APPID: *N3.xxxxxxx.181220150049
AUTHID : PSYSOP01 HOSTNAME: xxxxxxxx
EDUID : 65329 EDUNAME: db2agntp (XXXXXXXX) 1
FUNCTION: DB2 UDB, database utilities,
sqlubWriteEndSDBackupLogRecord, probe:374
MESSAGE : ZRC=0x82100001=-2112880639=SQLP_NONSEVERE_PRGERR
"DPS detects non-severe programming error"
DIA8532C An internal processing error has occurred.
DATA #1 : String, 63 bytes
Log file registration pending preventing backup from proceeding
DATA #2 : String, 30 bytes
sqlubWriteEndSDBackupLogRecord
CALLSTCK: (Static functions may not be resolved correctly, as
they are resolved to the nearest symbol)
[0] 0x0900000007FB9940 sqlzSetAndLog901 + 0x2C0
[1] 0x090000000DC735F0
sqlubWriteEndSDBackupLogRecord__FP8sqlrr_cbP23SQLKD_RQST_REPLY_F
ORMATUlPP15SQLR_RPCMESSAGE + 0xBF0
[2] 0x0900000008AAF200 sqlrr_rpc_router__FP8sqlrr_cb + 0x13C0
[3] 0x0900000008AAC280
sqlrr_subagent_router__FP8sqeAgentP12SQLE_DB2RA_T + 0xF80
[4] 0x0900000008A91420
IPRA.$sqleSubRequestRouter__FP8sqeAgentPUiPUl + 0x7C0
[5] 0x0900000008A8B4C8 sqleProcessSubRequest__FP8sqeAgent +
0x2C28
[6] 0x090000000735A02C RunEDU__8sqeAgentFv + 0x9EC
[7] 0x090000000733CD20 EDUDriver__9sqzEDUObjFv + 0x2E0
[8] 0x090000000722BC14 sqloEDUEntry + 0x374
[9] 0x09000000009D5FE8 _pthread_body + 0xE8
[10] 0xFFFFFFFFFFFFFFFC ?unknown + 0xFFFFFFFF

2018-12-21-00.02.46.826992+540 I5237A617 LEVEL: Error
PID : 37093746 TID : 65329 PROC : db2sysc 1
INSTANCE: xxxxxxx NODE : 001 DB : XXXXXXXX
APPHDL : 3-276 APPID: *N3.xxxxxxx.181220150049
AUTHID : PSYSOP01 HOSTNAME: xxxxxxxx
EDUID : 65329 EDUNAME: db2agntp (XXXXXXXX) 1
FUNCTION: DB2 UDB, database utilities,
sqlubWriteEndSDBackupLogRecord, probe:374
DATA #1 : SQLPG_EXTENT_NUM, PD_TYPE_SQLPG_EXTENT_NUM, 4 bytes
53
DATA #2 : unsigned integer, 8 bytes
3539402753
DATA #3 : unsigned integer, 8 bytes
3539402798

****************************************************************
* USERS AFFECTED:                                              *
* ALL                                                          *
****************************************************************
* PROBLEM DESCRIPTION:                                         *
* See Error Description                                        *
****************************************************************
* RECOMMENDATION:                                              *
* Upgrade to Db2 11.1 Mod 4 Fixpack 5 or higher                *
****************************************************************

-Before issue the backup, customer can run a small transaction
(eg. insert one row into a dummy table) from every member.  This
should avoid the boundary condition that triggers the problem.

-The following query can be used to detect boundary condition:

    select MEMBER, LOG_WRITES from
table(mon_get_transaction_log(-2))

If any member returned a value of 0 for LOG_WRITES, it means
that member has not written any log records to disk (so likely
no update transaction has executed on that member).  If ALL
member return non-zero value, BACKUP command should not run into
this issue.

not known / see Local fix

Date  - problem reported    :
Date  - problem closed      :
Date  - last modified       :

05.02.2019
16.01.2020
16.01.2020