The restore database command could fail and it could cause the
instance crash in the following condition.
1. Backup command was executed with TO option with multiple
directories.
2. Execute restore database command against the backup image
taken above and the command has HISTORY option and FROM option
with one directory.
Example:
* Executing backup command having TO option with three
directories.
$ db2 "BACKUP DATABASE DBNAME TO
'/home/db2inst/backup','/home/db2inst/backup','/home/db2inst/bac
kup' WITH 4 BUFFERS"
Backup successful. The timestamp for this backup image is :
20160731225703
* Executing restore database command for the above backup
command having FROM option with onde directory.
$ db2 "restore db DBNAME history file from
'/home/db2inst/backup' taken at 20160731225703"
SQL1224N The database manager is not able to accept new
requests, has
terminated all requests in progress, or has terminated the
specified request
because of an error or a forced interrupt. SQLSTATE=00000
At this error, the instance crashed with the following
db2diag.log
2016-07-31-22.57.53.582793-240 E37600A425 LEVEL: Severe
PID : 29032624 TID : 1946 PROC : db2sysc
INSTANCE: db2inst NODE : 000 DB : DBNAME
APPHDL : 0-35 APPID:
*LOCAL.db2inst.160801025753
AUTHID : DB2INST
EDUID : 1946 EDUNAME: db2agent (DBNAME)
FUNCTION: DB2 UDB, database utilities, sqludrsa, probe:800
MESSAGE : Restore Terminated.
<snippet>
2016-07-31-22.59.09.189955-240 I44738A560 LEVEL: Severe
PID : 29032624 TID : 2002 PROC : db2sysc
INSTANCE: db2inst NODE : 000 DB : DBNAME
APPHDL : 0-38 APPID:
*LOCAL.db2inst.160801025909
AUTHID : DB2INST
EDUID : 2002 EDUNAME: db2agent (DBNAME)
FUNCTION: DB2 UDB, RAS/PD component,
pdResilienceIsSafeToSustain, probe:800
DATA #1 : String, 37 bytes
Trap Sustainability Criteria Checking
DATA #2 : Hex integer, 8 bytes
0x0000000501001800
DATA #3 : Boolean, 1 bytes
false
Note that this problem happens for V9.7 FP9 and later. |
