The hang edu db2loggw stacktrace is not complete:
Stack:
======================
0x09000000004D8570 _p_nsleep + 0x10
0x090000000002B744 nsleep + 0xE4
0x09000000001555A8 nanosleep + 0x188
0x090000000F367808 ossSleep + 0xA8
0x090000001FDAA16C sqlorest + 0x188
0x090000001D729930 sqlpgStartMfhEntry__FP9SQLP_DBCBCUiCUl +
0x8B8
But considering below message:
2014-03-01-12.44.43.531921-300 I36007A436 LEVEL: Info
PID : 11206932 TID : 12338 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE<<<<<
HOSTNAME: host1
EDUID : 12338 EDUNAME: db2loggr (SAMPLE) 0
FUNCTION: DB2 UDB, data protection services,
sqlpgSwitchProcessAllLogStreams, probe:1600 //the switch phase
DATA #1 : <preformatted>
Log stream 0, TailIndex 59, extNum 303894
2014-03-01-12.44.43.573499-300 I36444A576 LEVEL: Info
PID : 11206932 TID : 13109 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE<<<<<
HOSTNAME: host1
EDUID : 13109 EDUNAME: db2loggw (SAMPLE) 0
FUNCTION: DB2 UDB, data protection services, sqlpgOpenLogExtent,
probe:850
MESSAGE : ZRC=0x860F000A=-2045837302=SQLO_FNEX "File not found."
DIA8411C A file "" could not be found.
DATA #1 : <preformatted>
Error opening database log
/db2/SAMPLE/log_dir2/NODE0000/LOGSTREAM0000/S0303894.LOG.
sqlpgSwitchProcessAllLogStreams, probe:1600 means the redo phase
is nearly completed, db2 is performing a 'switch' phase by loggr
EDU. Before redo completes, there is only 1 place that loggw EDU
could be involved, which is sqlpgCrashRecoveryPostRedo.
loggw may need to write something into the log file, but it
failed to write to mirror log, so it called
sqlpgStartOrCompleteMfhEntryAndUpdateMFHIfNecessary to mark
mirror log path BAD. then it found dual array is full:
2014-03-01-12.44.43.592614-300 I37021A504 LEVEL: Info
PID : 11206932 TID : 13109 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE<<<<<
HOSTNAME: host1
EDUID : 13109 EDUNAME: db2loggw (SAMPLE) 0
FUNCTION: DB2 UDB, data protection services, sqlpgStartMfhEntry,
probe:7240
DATA #1 : <preformatted>
Dual log array is full. Checking if any new logs have been
archived.
end: 366957, lowestUnarchivedExtNum: 303893, path: 9
2014-03-01-12.44.43.594471-300 I37526A409 LEVEL: Info
PID : 11206932 TID : 13109 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE<<<<<
HOSTNAME: host1
EDUID : 13109 EDUNAME: db2loggw (SAMPLE) 0
FUNCTION: DB2 UDB, data protection services, sqlpgStartMfhEntry,
probe:7280
MESSAGE : Wait 5 minutes for log file to be archived
2014-03-01-12.44.43.595863-300 I37936A171 LEVEL: Warning
PID:11206932 TID:13109 NODE:000 Title: SQLP_LFH
Dump File:/db2/db2inst1/db2dump/11206932.13109.000.dump.bin
This APAR is to add sanity checking and also to "correct" any
dual log entries that we can determine are correct and that we
can determine an appropriate value to patch it to. |
