Hi

OS: LInux/Unix
DB2 : V 8.1

Often the db2inst1 instance goes down,then i need to issue db2start command to up the instance.

what could be the actual reason behind this one.

please let me know,It's urgent


Regards,

Sampath
DBA

I would suggest that you look in the db2diag.log file. This should tell you why it is failing.

Andy

Hi

These are some i got from db2diag.log


AddRemoveString CLASSPATH ${INSTHOME?}/sqllib/function a
2007-04-30-18.14.11.280762+180 I776038G499 LEVEL: Severe
PID : 15357 TID : 3086960320 PROC : db2agent (BHEBBS) 0
INSTANCE: db2inst1 NODE : 000 DB : BHEBBS
APPHDL : 0-15 APPID: *LOCAL.db2inst1.070430151402
FUNCTION: DB2 UDB, relation data serv, sqlrr_dump_ffdc, probe:902
RETCODE : ZRC=0x87120007=-2028863481=SQLR_SEVERE_PGM_ERROR
"Severe internal error"
DIA8516C A severe internal processing error has occurred.



2007-04-30-18.14.11.281019+180 I776538G338 LEVEL: Severe
PID : 15357 TID : 3086960320 PROC : db2agent (BHEBBS) 0
INSTANCE: db2inst1 NODE : 000 DB : BHEBBS
APPHDL : 0-15 APPID: *LOCAL.db2inst1.070430151402
FUNCTION: DB2 UDB, relation data serv, sqlrr_dump_ffdc, probe:902



Note:

I am able to connect as db2inst1 user,by giving the db2start command,

But when another user(db2rpt ) tries to connect it is showing (ODBC--failed).
so,the user it not able to connect .



Regards,

Sampath
DBA

You should be opening a PMR with IBM DB2 support when you get "severe internal errors".

Andy

yes, because the probe=902 indicates the action to be performed and not correctly accepted by the dbm - hence the first failure data capture (ffdc)