DB2 - Problembeschreibung
Problem IT05545 | Status: Geschlossen |
HADR RESOURCE STATE REPORTS AS OFFLINE ON BOTH NODES EVEN THOUGH THE HADR DATABASE IS IN PRIMARY STANDBY PEER STATE | |
Produkt: | |
DB2 FOR LUW / DB2FORLUW / A10 - DB2 | |
Problembeschreibung: | |
In a HADR/TSA environment where the DB2 instance name is a substring of the "db2V10_start.ksh" string, e.g. instance name = "db2", then the HADR resource reports as offline on both nodes. This is due to the following line within the hadrV10_monitor.ksh script: isInstanceStartRunning=$(ps -ef | grep db2V10_start.ksh | grep ${candidate_P_instance?} | wc -l) The above checks for the existence of the db2V10_start.ksh process. In the case where the DB2 instance name is a substring of "db2V10_start.ksh" then the following result is returned: $ ps -ef | grep db2V10_start.ksh | grep db2 db2inst1 17161 14568 0 15:32 pts/17 00:00:00 grep db2V10_start.ksh The hadrV10_monitor.ksh script interprets this result as if the db2V10_start.ksh process is currently running and returns 2(Offline). In order to resolve this issue, the "grep db2V10_start.ksh" process must be filtered out from the results. Making the following change to the above line from the script fixes this: isInstanceStartRunning=$(ps -ef | grep db2V10_start.ksh | grep -v grep |grep ${candidate_P_instance?} | wc -l) | |
Problem-Zusammenfassung: | |
**************************************************************** * USERS AFFECTED: * * DB2 HADR/HA users * **************************************************************** * PROBLEM DESCRIPTION: * * See Error Description * **************************************************************** * RECOMMENDATION: * * Upgrade to DB2 version 10.1.0.5 * **************************************************************** | |
Local-Fix: | |
Make the following change to the /usr/sbin/rsct/sapolicies/db2/hadrV10_monitor.ksh script on both nodes: Replace this line: isInstanceStartRunning=$(ps -ef | grep db2V10_start.ksh | grep ${candidate_P_instance?} | wc -l) with this line: isInstanceStartRunning=$(ps -ef | grep db2V10_start.ksh | grep -v grep | grep ${candidate_P_instance?} | wc -l) | |
Lösung | |
Fixed in DB2 version 10.1.0.5 | |
Workaround | |
keiner bekannt / siehe Local-Fix | |
Weitere Daten | |
Datum - Problem gemeldet : Datum - Problem geschlossen : Datum - der letzten Änderung: | 13.11.2014 14.07.2015 14.07.2015 |
Problem behoben ab folgender Versionen (IBM BugInfos) | |
Problem behoben lt. FixList in der Version | |
10.1.0.5 |