|
Problem IT08871
|
Status: Closed |
DB2READLOG API REPORTING SQL2654W INSTEAD OF SQL1273N FOR MISSING LOG ERROR CAUSING INCORRECT RESULT IN REPLICATION |
| product: |
DB2 FOR LUW / DB2FORLUW / A10 - DB2 |
| Problem description: |
This APAR occurs when the user has missing logs and utilized
db2ReadLog API to perform replication. The source of missing log
error is independent of this APAR. An example is the user
restored the database to an older time and started a new chain
of workload but did not save all the log files.
db2ReadLog API has only returned SQL2654W to indicate that all
active log have been read by the asynchronous reader instead of
reporting the missing log error. Replication continues with no
knowledge of the missing log error and eventually fails with
data inconsistency or data loss error.
When checking the db2diag.log, the following message can be seen
repeatedly:
2015-05-06-21.42.34.355129-240 I1891946E2752 LEVEL:
Warning
PID : 20646 TID : 46913986488640 KTID :
23576
PROC : db2sysc 0
INSTANCE: tester NODE : 000 DB :
TESTDB
APPHDL : 0-104 APPID: *N0.tester.150507014233
AUTHID : TESTER HOSTNAME: host1
EDUID : 138 EDUNAME: db2agent (TESTDB) 0
FUNCTION: DB2 UDB, recovery manager,
sqlpshrValidateLogStreamEndPoint, probe:1122
DATA #1 : <preformatted>
Some log records might not be returned on this call,will be
returned on next call.please collect the diag data and send it
to IBM support
CALLSTCK: (Static functions may not be resolved correctly, as
they are resolved to the nearest symbol)
[0] 0x00002AAAACD986AE pdLogVPrintf + 0x1B8
[1] 0x00002AAAACD98AB3 pdLogPrintf + 0xF3
[2] 0x00002AAAB0EAF536
_Z32sqlpshrValidateLogStreamEndPointP14sqlpMasterDbcbP17SQLPSHR_
MASTER_CBP17SQLPSHR_WORKER_CBmmbb + 0xDE8
[3] 0x00002AAAB0EB19FA
_Z36sqlpshrWaitForWorkerToQueueLogRecordP8sqeAgentP14sqlpMasterD
bcbP17SQLPSHR_MASTER_CBP17SQLPSHR_WORKER_CBPmPP21SQLPSHR_SORTED
+ 0x482
[4] 0x00002AAAB0EB2BFD
_Z15sqlpshrScanNextPvP8sqeAgentmPP16SQLPR_LOGREC_DISPm + 0x47B
[5] 0x00002AAAB0BDC9BA
_Z22sqlpgReadLogReadActionP8sqeAgentP14sqlpMasterDbcbjP15SQLP_RE
ADLOG_CBP17SQLP_READLOG_INFOmPc + 0x52C
[6] 0x00002AAAB0BE0E47
_Z15sqlpgReadLogAPIP8sqeAgentjP6db2LRIS2_S2_mmPcP17SQLP_READLOG_
INFOP5sqlca + 0xD0F
[7] 0x00002AAAB0F6D52B
_Z22sqlpReadLogInternalAPIP8sqeAgentjP6db2LRIS2_S2_PcjP17SQLP_RE
ADLOG_INFOjP5sqlca + 0x19D
[8] 0x00002AAAB0F6E3E8
_Z15sqlpReadLogDRDAP5sqldaS0_P8sqeAgentP5sqlca + 0xC22
[9] 0x00002AAAAE41503C
_Z19sqlerKnownProcedureiPcPiP5sqldaS2_P13sqlerFmpTableP8sqeAgent
P5sqlca + 0x1918
[10] 0x00002AAAAE41C512
_Z11sqlerCallDLP14db2UCinterfaceP9UCstpInfo + 0x24DE
[11] 0x00002AAAAE6FEAAF
_Z19sqljs_ddm_excsqlsttP14db2UCinterfaceP13sqljDDMObject + 0x4EB
[12] 0x00002AAAAE6CDBE2
_Z21sqljsParseRdbAccessedP13sqljsDrdaAsCbP13sqljDDMObjectP14db2U
Cinterface + 0x25C
[13] 0x00002AAAAE6CEE69
_Z10sqljsParseP13sqljsDrdaAsCbP14db2UCinterfaceP8sqeAgentb +
0x239
[14] 0x00002AAAAE6C1427
/home/tester/sqllib/lib64/libdb2e.so.1 + 0x39E0427
[15] 0x00002AAAAE6C363F
/home/tester/sqllib/lib64/libdb2e.so.1 + 0x39E263F
[16] 0x00002AAAAE6BEAEF
/home/tester/sqllib/lib64/libdb2e.so.1 + 0x39DDAEF
[17] 0x00002AAAAE6BE35C
_Z17sqljsDrdaAsDriverP18SQLCC_INITSTRUCT_T + 0x29A
[18] 0x00002AAAADE060B2 _ZN8sqeAgent6RunEDUEv + 0x4F8
[19] 0x00002AAAB324B40C _ZN9sqzEDUObj9EDUDriverEv + 0x17C
[20] 0x00002AAAB324B034 _Z10sqlzRunEDUPcj + 0x26
[21] 0x00002AAAB09FAAAF sqloEDUEntry + 0x971
[22] 0x00002AAAAABCE2A3 /lib64/libpthread.so.0 + 0x62A3
[23] 0x00002AAAB988E14D __clone + 0x6D
2015-05-06-21.42.34.370459-240 I1894699E200 LEVEL: Warning
PID:20646 TID:46913986488640 NODE:000 Title: SQLPSHR_WORKER_CB
Dump
File:/home/tester/sqllib/db2dump/DIAG0000/20646.138.000.dump.bin
The probe point 1122 indicates that the missing log error is
masked out and we have dumped diagnostic information instead. |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* db2ReadLog API users in all environment *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* V101 users are recommended to upgarde to fixpack FP5. V105 *
* users are recommended to wait for the patch in FP7. *
**************************************************************** |
| Local Fix: |
Reinitiate replication by refreshing all the data from the
replication source |
| Solution |
Problem is first fixed in V101 FP5 and V105 FP7 for purescale
users because this leads to data corruption.
For ESE users, the existing behaviour is kept (diagnostic
information is dumped out in flight recorder and db2ReadLog API
doesn't return the missing log error) to avoid false alarms. |
| Workaround |
Reinitiate replication by refreshing all the data from the
replication source |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 12.05.2015
20.07.2015
20.07.2015 |
| Problem solved at the following versions (IBM BugInfos) |
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 10.1.0.5
|
 |
