Environment: DB2 V9.7 FP 3a, Multipartitioned environment with 17 partitions spanning across 3 physical nodes
Operating System: Linux

Backup failed with SQL02428N.

We have an online backup with include logs. I see a strange situation here.Backup has failed complaining that it couldn't retrieve a logfile. But the strange thing is that the logfile which it is complaining was already archived 12 hours before the backup started.



Take the example for NODE 1 in the below scenario. The log file would be S0016416.LOG. Notice the timestamps



Here is the scenario:



DBPARTITIONNUM START_TIME END_TIME FIRSTLOG LASTLOG OPERATION SQLCODE SQLSTATE

-------------- -------------- -------------- --------------- --------------- --------- ----------- --------

0 20110713001815 20110713013940 S0030994.LOG S0030995.LOG B -2001

1 20110713001815 20110713013905 S0016416.LOG S0016417.LOG B -2428

2 20110713001815 20110713013858 S0016118.LOG S0016119.LOG B -2428

3 20110713001815 20110713013906 S0013629.LOG S0013630.LOG B -2428

4 20110713001815 20110713014046 S0020834.LOG S0020834.LOG B -2001

5 20110713001815 20110713014047 S0015487.LOG S0015487.LOG B -2001

6 20110713001815 20110713014042 S0015853.LOG S0015853.LOG B -2001

7 20110713001815 20110713013905 S0013972.LOG S0013973.LOG B -2428

8 20110713001815 20110713014056 S0020295.LOG S0020295.LOG B -2001

9 20110713001815 20110713013928 S0016893.LOG S0016894.LOG B -2428

10 20110713001815 20110713013939 S0015917.LOG S0015918.LOG B -2428

11 20110713001815 20110713013928 S0016894.LOG S0016895.LOG B -2428

12 20110713001815 20110713013928 S0016929.LOG S0016930.LOG B -2428

13 20110713001815 20110713013928 S0016011.LOG S0016012.LOG B -2428

14 20110713001815 20110713014021 S0016719.LOG S0016719.LOG B -2001

15 20110713001815 20110713013928 S0020168.LOG S0020169.LOG B -2428

16 20110713001815 20110713013928 S0016153.LOG S0016154.LOG B -2428




Here is the error message:

2011-07-13-01.37.46.101342-240 E27402310E568 LEVEL: Severe

PID : 15386 TID : 46915601295680PROC : db2sysc 1

INSTANCE: lrmp NODE : 001 DB : LRMDB

APPHDL : 0-9467 APPID: *N0.lrmp.110713041815

AUTHID : LRMP

EDUID : 42339 EDUNAME: db2bm.8879.1 (LRMDB) 1

FUNCTION: DB2 UDB, database utilities, sqlubProcessLogExtent, probe:1790

MESSAGE : ADM8010E Backup was unable to copy requested log file "S0016416.LOG"

for inclusion in the backup image. The backup has been aborted.







2011-07-12-11.57.12.334404-240 E27021344E472 LEVEL: Info

PID : 15386 TID : 46912891775296PROC : db2sysc 1

INSTANCE: lrmp NODE : 001

EDUID : 111 EDUNAME: db2logmgr (LRMDB) 1

FUNCTION: DB2 UDB, data protection services, sqlpgArchiveLogFile, probe:3180

DATA #1 : <preformatted>

Completed archive for log file S0016416.LOG to /db2logs/lrmp/archive/lrmp/LRMDB/NODE0001/C0000006/ from /db2logs/lrmp/active/NODE0001/.

Logs are archived as they become full; they can remain active until transactions stop using them. So, S0016416.LOG became full and was archived at 2011-07-12-11.57.12.334404, although the transaction that had been using it remained active until at least 2011-07-13-01.37.46.101342.

Is there any information in the diag.log as to why it couldn't retrieve this log?



Do you have some script/process that moves the logs from the archived location - /db2logs/lrmp/archive/lrmp/LRMDB/NODE0001/C0000006/ ?

@n_i, I guess the transaction using log 16416 was active at 2011-07-13-00.18.15 but was committed/rolled back before 2011-07-13-01.37.46.101342. If the transaction had been active at 1.37.46, the log would have been available in the active log directory.

@Bella : It couldn't be retrieved because. We have a process which moves the archived logs to net backups and deletes the log file from active log directory.