hi

database was marked bad log from notify file says

ADM1532E Crash recovery has failed with SQLCODE "-1042".
▲
2009-12-01-12.36.02.716636 Instance:VATPROD Node:000
PID:47216(db2agent (VATPROD) 0) TID:1 Appid:0A9B0126.09F8.091209122746
base sys utilities sqleMarkDBad Probe:10 Database:VATPROD

ADM7518C "VATPROD" marked bad.

server was rebooted and when db was started it started sucessfully when tried connecting error first it gave a I/O error then
its started giving
SQL1042C An unexpected system error occurred. SQLSTATE=58004
tried db2stop and db2start, tried restarting db , tried deactivation and activation, no results , only options to restore wht to know what could have gone wrong , could be some file system must have corrupted
regds
Paul

sql1042c is very generic, you need to check db2diag.log to see why crash recovery is failing. Rename your existing db2diag.log (you can use "db2diag -A" command if you like) and connect to the db. If connection fails, you'll see messages about crash recovery being started, using so many agents... and then you'll see the errors (ie. bad page, bad log, etc...). Also, check the system logs for any errors.

when u try connecting to db error was updating in notify file and db2diag was not giving any info
"ADM1532E Crash recovery has failed with SQLCODE "-1042".
▲
2009-12-01-12.36.02.716636 Instance:VATPROD Node:000
PID:47216(db2agent (VATPROD) 0) TID:1 Appid:0A9B0126.09F8.091209122746
base sys utilities sqleMarkDBad Probe:10 Database:VATPROD

ADM7518C "VATPROD" marked bad"
--
i doubt on this what happen, as system guy informed tht he has zip the transaction log files for space in the mount point, could this be a reason if it was trying for crash recovery it was not able to find the log files and recovery stopped and db was marked bad, but it should give an error saying not able find the logs or something ??
regds
Paul

did you unzip the logs and try again ????

Check what the diaglevel is set to.

sathya point to think yes unziped the log file still crash recovery failed and gave the same error,,
bella
diag is set to 3
regds
Paul

try it with 4 ... that may throw more info .

Have you checked the syslog ... Sometimes you get useful information there

Also, check permissions on the db2diag.log - it should be owned by the instance id and at least has rw for the owner. I suppose permissions on the dir is ok if db2 can update the notify log. You should see a lot of information logged with diaglevel 3.


It may also be a good idea to check the ownership/permissions on the log files and make sure db2 can find them in the active log dir (if the problem is related to db2 being unable to find the logs during crash recovery).

thanx guys

backup was restored it was development server, so its ok.

regds
Paul