DB2 FOR LUW / DB2FORLUW / 970 - DB2

** This APAR applies only to integrated HA solutions ** 
 
On memory constrained systems or very busy systems ps behavior 
is unpredictable where ps may return the process name in square 
brackets. 
Hence there's a chance that this check: 
 
p_pid=$(ps -u ${candidate_P_instance?} -o args | grep -c 
"^db2sysc") 
 
returns 0, which in turns makes hadrVxx_monitor.ksh script 
return a status of 2 i.e. HADR resource is down. 
 
Additionally the following line with the p_out parameter ends up 
storing all the ps output on one line, which is then grepped. 
When that line contains a db2sysc entry in brackets, it returns 
with a 0 return code which places HADR in an unknown state: 
 
p_out=$(ps -u ${candidate_P_instance?} -o args | egrep 
"^db2sysc|^\[db2sysc\]"| grep -v defunct)

**************************************************************** 
* USERS AFFECTED:                                              * 
* All using HADR TSA scripts.                                  * 
**************************************************************** 
* PROBLEM DESCRIPTION:                                         * 
* See Error Description                                        * 
**************************************************************** 
* RECOMMENDATION:                                              * 
* Upgrade to DB2 v9.7 FP9.                                     * 
****************************************************************

To avoid these "false negatives", the above lines have been 
modified as follows: 
 
OLD 
 
   # We need to checking if the instance start is complete here. 
To minimize 
   # the dependency on the checking for the db2 start script, we 
first check 
   # if there is a db2sysc process. If not, no need to go 
further. If there is 
   # a db2sysc process, check if the script is still running. 
   p_pid=$(ps -u ${candidate_P_instance?} -o args | grep -c 
"^db2sysc") 
   if [ $p_pid -eq 0 ]; then 
      logger -i -p info -t $PROGNAME "$0 $DB2INSTANCE is 
offline, hadr monitor must return offline" 
      rc=2 
      return $rc 
   elif [ $p_pid -ne 0 ]; then 
      isInstanceStartRunning=$(ps -ef | grep db2V97_start.ksh | 
grep ${candidate_P_instance?} | wc -l) 
      if [[ $isInstanceStartRunning -ne 0 ]]; then 
         logger -i -p notice -t $0 "db2sysc started but Instance 
start still running...return OFFLINE for HADR database 
${DB2HADRDBNAME?}" 
         rc=2 
         return $rc 
      fi 
   fi 
 
 
NEW 
   # We need to checking if the instance start is complete here. 
To minimize 
   # the dependency on the checking for the db2 start script, we 
first check 
   # if there is a db2sysc process. If not, no need to go 
further. If there is 
   # a db2sysc process, check if the script is still running. 
   p_out=$(ps -u ${candidate_P_instance?} -o args | egrep 
"^db2sysc|^\[db2sysc\]"| grep -v defunct) 
   p_pid=$(echo $p_out | grep -v "\[db2sysc\]" | grep -c 
"db2sysc") 
   if [ $p_pid -eq 0 ]; then 
      p_pid=$(echo $p_out | grep -c "\[db2sysc\]") 
      if [[ $p_pid != 0 ]]; then 
         log err "ps returns only [db2sysc]: returning 0" 
         rc=0 
      else 
         logger -i -p info -t $PROGNAME "$0 $DB2INSTANCE is 
offline, hadr monitor must return offline" 
         rc=2 
      fi 
      return $rc 
   elif [ $p_pid -ne 0 ]; then 
      isInstanceStartRunning=$(ps -ef | grep db2V97_start.ksh | 
grep ${candidate_P_instance?} | wc -l) 
      if [[ $isInstanceStartRunning -ne 0 ]]; then 
         logger -i -p notice -t $0 "db2sysc started but Instance 
start still running...return OFFLINE for HADR database 
${DB2HADRDBNAME?}" 
         rc=2 
         return $rc 
      fi 
   fi 
 
 
OLD 
 
 p_out=$(ps -u ${candidate_P_instance?} -o args | egrep 
"^db2sysc|^\[db2sysc\]"| grep -v defunct) 
   p_pid=$(echo $p_out | grep -v "\[db2sysc\]" | grep -c 
"db2sysc") 
   if [ $p_pid -eq 0 ]; then 
      p_pid=$(echo $p_out | grep -c "\[db2sysc\]") 
      if [[ $p_pid != 0 ]]; then 
         log err "ps returns only [db2sysc]: returning 0" 
         rc=0 
 
 
NEW 
 
   p_out=$(ps -u ${candidate_P_instance?} -o args | egrep 
"^db2sysc|^\[db2sysc\]"| grep -v defunct) 
   p_pid=$(echo $p_out | tr " " "\n" | grep -v "\[db2sysc\]" | 
grep -c "db2sysc") 
   if [ $p_pid -eq 0 ]; then 
      p_pid=$(echo $p_out | grep -c "\[db2sysc\]") 
      if [[ $p_pid != 0 ]]; then 
         log err "ps returns only [db2sysc]: returning 0" 
         rc=0

DB2 Version 9.7 Fix Pack 9 for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 9a for Linux, UNIX, and Windows
 DB2 Version 9.7 Fix Pack 10 for Linux, UNIX, and Windows
 

Fixed in DB2 v9.7 FP9.

not known / see Local fix

forerunner  : APAR is sysrouted TO one or more of the following: IC97638 IC97667 
follow-up : 

Date  - problem reported    :
Date  - problem closed      :
Date  - last modified       :

01.08.2013
23.12.2013
23.12.2013

9.7.FP9