If there is a load copy yes into /dev/null, the subsequent
rollforward to restore the load copy would fail with sql3785N
error due to the special device /dev/null used for the load
copy. However after the failure, the tablespace may wrongly be
left in the normal state. In some situations, this may lead the
further rollfoward of the log records for the same table to end
up with corruption errors.
The rollforward should fail and put the tablespace into restore
pending state.
For Example:
2013-08-13-16.52.20.364464+600 I1817A467 LEVEL: Warning
PID : 39977030 TID : 7858 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000 DB : HADRDB1
APPHDL : 0-5742 APPID: *LOCAL.DB2.130813064550
EDUID : 7858 EDUNAME: db2agent (HADRDB1) 0
FUNCTION: DB2 UDB, database utilities, sqludcpy, probe:553
DATA #1 : String, 60 bytes
Starting to restore a load copy.
DB2INST9.MYT.20130813165219
2013-08-13-16.52.20.369275+600 E2285A397 LEVEL: Error
PID : 39977030 TID : 9665 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000
EDUID : 9665 EDUNAME: db2med.7858.1 0
FUNCTION: DB2 UDB, database utilities, sqluMCInitRestoreMC,
probe:1159
MESSAGE : Media controller -- Unable to initialize device
/dev/null (zrc =
0x00000000)
2013-08-13-16.52.20.369527+600 E2683A360 LEVEL: Error
PID : 39977030 TID : 9665 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000
EDUID : 9665 EDUNAME: db2med.7858.1 0
FUNCTION: DB2 UDB, database utilities, sqluMCInitRestoreMC,
probe:1159
MESSAGE : SQL2061N An attempt to access media "" is denied.
2013-08-13-16.52.20.369866+600 E3044A335 LEVEL: Error
PID : 39977030 TID : 9665 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000
EDUID : 9665 EDUNAME: db2med.7858.1 0
FUNCTION: DB2 UDB, database utilities, sqluMCInitRestoreMC,
probe:1159
DATA #1 : String, 9 bytes
/dev/null
2013-08-13-16.52.20.370126+600 E3380A376 LEVEL: Severe
PID : 39977030 TID : 9665 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000
EDUID : 9665 EDUNAME: db2med.7858.1 0
FUNCTION: DB2 UDB, database utilities,
sqluMCProcessRestoreStates, probe:786
MESSAGE : Unexpected media type.
DATA #1 : signed integer, 1 bytes
78
<snippet>
2013-08-13-17.03.05.789904+600 I54604A420 LEVEL: Warning
PID : 39977030 TID : 7858 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000 DB : HADRDB1
APPHDL : 0-5742 APPID: *LOCAL.DB2.130813064550
EDUID : 7858 EDUNAME: db2agent (HADRDB1) 0
FUNCTION: DB2 UDB, database utilities, sqludcpy, probe:1130
MESSAGE : Load copy restore terminated.
2013-08-13-17.03.05.790091+600 I55025A443 LEVEL: Error
PID : 39977030 TID : 7858 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000 DB : HADRDB1
APPHDL : 0-5742 APPID: *LOCAL.DB2.130813064550
EDUID : 7858 EDUNAME: db2agent (HADRDB1) 0
FUNCTION: DB2 UDB, recovery manager, sqlpReplayMaster,
probe:1100
MESSAGE : Load copy image restore terminated with rc -3785
2013-08-13-17.03.05.824500+600 I55469A444 LEVEL: Warning
PID : 39977030 TID : 8380 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000 DB : HADRDB1
APPHDL : 0-5742 APPID: *LOCAL.DB2.130813064550
EDUID : 8380 EDUNAME: db2redom (HADRDB1) 0
FUNCTION: DB2 UDB, recovery manager, sqlpRFWppl, probe:870
MESSAGE : Tablespaces taken offline due to failed load recovery.
2013-08-13-17.04.12.677465+600 I55914A444 LEVEL: Warning
PID : 39977030 TID : 8380 PROC : db2sysc
0
INSTANCE: db2ins18 NODE : 000 DB : HADRDB1
APPHDL : 0-5742 APPID: *LOCAL.DB2.130813064550
EDUID : 8380 EDUNAME: db2redom (HADRDB1) 0
FUNCTION: DB2 UDB, recovery manager, sqlpRFWppl, probe:870
MESSAGE : Tablespaces taken offline due to failed load recovery.
However, db2pd -tablespace output shows the tablespace is still
in normal state after the failure:
Address Id Type Content PageSz ExtentSz Auto
Prefetch BufID BufIDDisk FSC NumCntrs MaxStripe LastConsecPg
Name
0x0A0005000AED7280 12 SMS Regular 4096 32 No 32
1 1 Off 1 0 31 MYTBSPACE1
Address Id TotalPgs UsablePgs UsedPgs
PndFreePgs FreePgs HWM Max HWM State
MinRecTime NQuiescers PathsDropped
0x0A0005000AED7280 12 3738 3738 3738 0
0 0 0 0x00000000 1375141345 0
No |
