In the db2diag.log you will see messages like this:
2015-06-11-14.30.15.143760-300 I163341A515 LEVEL:
Warning
PID : 12976138 TID : 42664 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE
APPHDL : 0-12 APPID: *LOCAL.DB2.150607221510
HOSTNAME: SERVER02
EDUID : 42664 EDUNAME: db2redom (SAMPLE) 0
FUNCTION: DB2 UDB, recovery manager, sqlpPRecReadLog, probe:1435
MESSAGE : Almost too many logs
DATA #1 : db2LogStreamIDType, PD_TYPE_DB2_LOG_STREAM_ID, 2 bytes
0
2015-06-11-14.44.17.672907-300 I177884A1014 LEVEL: Error
PID : 12976138 TID : 42664 PROC :
db2sysc 0
INSTANCE: db2inst1 NODE : 000 DB :
SAMPLE
APPHDL : 0-12 APPID: *LOCAL.DB2.150607221510
HOSTNAME: SERVER02
EDUID : 42664 EDUNAME: db2redom (SAMPLE) 0
FUNCTION: DB2 UDB, recovery manager, sqlpPRecReadLog, probe:1434
MESSAGE : ZRC=0x8710001D=-2028994531=SQLP_LERR "Fatal Logic
Error"
DIA8526C A fatal error occurred in data protection
services.
DATA #1 : String, 14 bytes
Too many logs.
DATA #2 : String, 64 bytes
logStreamId / lowtranlsn / minbufflsn / minPseudoLsn / recordLsn
DATA #3 : db2LogStreamIDType, PD_TYPE_DB2_LOG_STREAM_ID, 2 bytes
0
DATA #4 : SQLP_LSN8, PD_TYPE_SQLP_LSN8, 8 bytes
000035705576C71F
DATA #5 : SQLP_LSN8, PD_TYPE_SQLP_LSN8, 8 bytes
00003570BDFB4858
DATA #6 : SQLP_LSN8, PD_TYPE_SQLP_LSN8, 8 bytes
FFFFFFFFFFFFFFFF
DATA #7 : SQLP_LSN8, PD_TYPE_SQLP_LSN8, 8 bytes
00003570BEE383EF
Then the database crashes. The cause is that the logs are not
being cleaned up on the standby, creating the "too many logs"
scenario. |
