用db2cos监控db2diag中的特定错误信息并抓取相应的app信息

Copy the db2cos script <instance home>/sqllib/bin/ to <instance home>/sqllib/adm/ and modify below section(uncomment one line of errstring as you need):
Modify below section:
   "DIAGSTRING")
      echo "Diag String Caught"                                      >> $logfile
      if [ ! -n "$database" ]
      then
         db2pd -inst                                                 >> $db2pd_outfile
      else
         diagfile=$diagpath\db2diag.log
         #get lastest MAX_LOG or NUM_LOG_SPAN or lock_escalation info from diag
         echo "-------info from diag" >> $logfile
         #MAX_LOG
         #errstring='ADM1542W'
         #NUM_LOG_SPAN
         #errstring='ADM1541W'
         #lock_escalation
         errstring='ADM5502W'
         grep -ip $errstring $diagfile|tail -11 >> $logfile
         #get application handle
         appnum=`echo $apphdl|awk -F '-' '{print $2}'`
         #get snapshot for this application
         echo "The application info which causing the error - APPID:$appnum" >> $logfile
         db2 get snapshot for application agentid $appnum >> $logfile
      fi
   ;;

Run command as below:
Run command to start catch info:
MAX_LOG:
db2pdcfg -catch diagstr='ADM1542W' count=1
NUM_LOG_SPAN:
db2pdcfg -catch diagstr='ADM1541W' count=1
lock escalation:
db2pdcfg -catch diagstr='ADM5502W' count=1

clear catch after things happened:
db2pdcfg -catch clear

check diagpath for file like *.cos.txt

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/75336/viewspace-2143115/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/75336/viewspace-2143115/