When online recovery feature is enabled (via DB2_ONLINERECOVERY
registry variable), the rolling back of inflight transaction
might be deferred during synchronous phase of database recovery.
Insufficient log space was reserved for the deferred
transaction, which can result in a failure to process the
deferred rollback (due to log space being consumed by new
transactions).
You may see db2diag.log message similar to these:
2018-12-18-16.52.58.264139-300 I282863E534 LEVEL:
Error
PID : 9937 TID : 140318821312256 PROC :
db2sysc
INSTANCE: db2ins NODE : 000 DB : X
APPHDL : 0-50 APPID: *LOCAL.DB2.181218215312
HOSTNAME: hotel
EDUID : 117 EDUNAME: db2agent (X)
FUNCTION: DB2 UDB, recovery manager, sqlprudm, probe:15550
MESSAGE : ZRC=0x85100009=-2062548983=SQLP_NOSPACE
"Log File has reached its saturation point"
DIA8309C Log file was full.
2018-12-18-16.52.58.264390-300 I283398E729 LEVEL:
Error
PID : 9937 TID : 140318821312256 PROC :
db2sysc
INSTANCE: db2ins NODE : 000 DB : X
APPHDL : 0-50 APPID: *LOCAL.DB2.181218215312
HOSTNAME: hotel
EDUID : 117 EDUNAME: db2agent (X)
FUNCTION: DB2 UDB, recovery manager, sqlprAsyncUndo, probe:1550
MESSAGE : ZRC=0x85100009=-2062548983=SQLP_NOSPACE
"Log File has reached its saturation point"
DIA8309C Log file was full.
DATA #1 : unsigned integer, 2 bytes
0
DATA #2 : unsigned integer, 2 bytes
0
DATA #3 : unsigned integer, 2 bytes
0
DATA #4 : unsigned integer, 2 bytes
0
DATA #5 : unsigned integer, 2 bytes
0
2018-12-18-16.52.58.264805-300 I284128E698 LEVEL:
Error
PID : 9937 TID : 140318821312256 PROC :
db2sysc
INSTANCE: db2ins NODE : 000 DB : X
APPHDL : 0-50 APPID: *LOCAL.DB2.181218215312
HOSTNAME: hotel
EDUID : 117 EDUNAME: db2agent (X)
FUNCTION: DB2 UDB, recovery manager, sqlprAsyncUndo, probe:27110
MESSAGE : ZRC=0x85100009=-2062548983=SQLP_NOSPACE
"Log File has reached its saturation point"
DIA8309C Log file was full.
DATA #1 :
Force shutdown of database due to error during rollback
processing. Asynchronous Undo will be disabled for the next
crash recovery.
If you see the message above, this indicates that a subsequent
attempt for crash recovery will not use the ONLINE RECOVERY
feature which will avoid this problem.
To prevent future occurrences of this problem, you can disable
the ONLINERECOVERY feature: db2set DB2_ONLINERECOVERY=0 |
