|
Problem IC99237
|
Status: Closed |
DEADLATCH BETWEEN INSTANCE QUIESCE AND HADR START DUE TO LATCHING ORDER |
| product: |
DB2 FOR LUW / DB2FORLUW / A50 - DB2 |
| Problem description: |
During an instance quiesce if an HADR start is issued, it can
result in a deadlatch between the quiesce and the agent that
invoked the HADR start. You will likely see the following
messages in the db2diag.log very close to each other in time:
2013-12-01-04.40.28.640481-300 E4372931E543 LEVEL:
Warning
PID : 12667 TID : 46912774334784PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
APPHDL : 0-617 APPID:
*LOCAL.db2inst1.131201093341
AUTHID : db2inst1
EDUID : 58 EDUNAME: db2agent (instance) 0
<-- EDU 58
FUNCTION: DB2 UDB, base sys utilities, sqleInstanceQuiesce,
probe:1
MESSAGE : ADM7500W A request was made to quiesce an instance
with the
following quiesce mode: "QUIESCE INSTANCE"
(...)
2013-12-01-04.40.34.308463-300 I4373900E469 LEVEL:
Warning
PID : 12667 TID : 46915962005824PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB : SAMPLE
APPHDL : 0-601 APPID:
*LOCAL.db2inst1.131201093738
AUTHID : db2user
EDUID : 272 EDUNAME: db2agent (SAMPLE) 0
<-- EDU 272
FUNCTION: DB2 UDB, High Availability Disaster Recovery,
hdrEduStartup, probe:211
51
MESSAGE : Info: HADR Startup has begun.
Collecting stacks on the above EDUIDs will show the following in
the latch section of the stack file:
12667.58.000.stack.txt <-- stack file for EDU 58
Latches:
Waiting on latch type: (SQLO_LT_sqeLocalDatabase__dblatch) -
Address: (0x200759dc8)
Holding Latch type: (SQLO_LT_sqeDBMgr__dbMgrLatch) - Address:
(0x2007501e8)
Stack Trace:
ossDumpStackTraceEx
OSSTrapFile::dumpEx
sqlo_trce + 0x03fb
sqloDumpDiagInfoHandler
sched_yield
sqloSpinLockConflict
sqeLocalDatabase::GetDbStateMap
sqleInstanceQuiesce
sqlerKnownProcedure
sqlerCallDL
sqljs_ddm_excsqlstt
sqljsParseRdbAccessed
sqljsParse
sqljsDrdaAsDriver
sqeAgent::RunEDU
sqzEDUObj::EDUDriver
sqlzRunEDU
sqloEDUEntry
12667.272.000.stack.txt <-- Stack file for EDU 272
Latches:
Waiting on latch type: (SQLO_LT_sqeDBMgr__dbMgrLatch) - Address:
(0x2007501e8)
Holding Latch type: (SQLO_LT_sqeLocalDatabase__dblatch) -
Address: (0x200759dc8)
Showing that each stack is holding the latch that the other is
waiting on |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* Any using the HADR feature *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* Upgrade to DB2 Cancun Release 10.5.0.4 (also known as Fix *
* Pack 4) for Linux, UNIX, and Windows *
**************************************************************** |
| Local Fix: |
n/a |
| available fix packs: |
DB2 Cancun Release 10.5.0.4 (also known as Fix Pack 4) for Linux, UNIX, and Windows
DB2 Version 10.5 Fix Pack 9 for Linux, UNIX, and Windows
|
| Solution |
First fixed in DB2 Cancun Release 10.5.0.4 (also known as Fix
Pack 4) for Linux, UNIX, and Windows |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 07.02.2014
15.09.2014
15.09.2014 |
| Problem solved at the following versions (IBM BugInfos) |
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 10.5.0.4
|
 |
