DB2 - Problembeschreibung
Problem IC65837 | Status: Geschlossen |
AFTER A TAKEOVER, HADR MONITOR SCRIPT MAY SHOW AN ONLINE HADR DATABASE ON THE OLD PRIMARY NODE. | |
Produkt: | |
DB2 FOR LUW / DB2FORLUW / 970 - DB2 | |
Problembeschreibung: | |
In an automated cluster controlled HADR environment, after a primary node reboot, there is a small timing hole where HADR monitor script may show that the HADR database resource online. In the OS log, you will see an instance start message after reboot of old primary node (amdha1). But before the instance start, HADR monitor script may return a return code of 1 i.e., online for HADR database resource. Jan 20 11:43:30 amdha1 db2V95_start.ksh[5104]: Entered /usr/sbin/rsct/sapolicies/db2/db2V95_start.ksh, db2inst5, 0 Jan 20 11:43:30 amdha1 /usr/sbin/rsct/sapolicies/db2/hadrV95_monitor.ksh[5123]: Returning 1 : db2inst5 db2inst5 UC4G Jan 20 11:43:39 amdha1 db2V95_start.ksh[5878]: Returning 0 from /usr/sbin/rsct/sapolicies/db2/db2V95_start.ksh ( db2inst5, 0) Due to this timing issue, TSAMP will see the same HADR database resource online on both servers. A resource cannot be online on more than one node concurrently in a TSAMP managed environment. So it will stop all instances of HADR database resource, and start it up on the old primary due to the order of the NodeNameList. This gives an appearance of an automatic failback after the original primary node rejoins. | |
Problem-Zusammenfassung: | |
**************************************************************** * USERS AFFECTED: * * Users with HADR on a cluster env * **************************************************************** * PROBLEM DESCRIPTION: * * AFTER A TAKEOVER, HADR MONITOR SCRIPT MAY SHOW AN * * ONLINEHADRDATABASE ON THE OLD PRIMARY NODE. * **************************************************************** * RECOMMENDATION: * * Upgrade to 9.7 FP2 * **************************************************************** | |
Local-Fix: | |
verfügbare FixPacks: | |
DB2 Version 9.7 Fix Pack 2 for Linux, UNIX, and Windows | |
Lösung | |
AFTER A TAKEOVER, HADR MONITOR SCRIPT MAY SHOW AN ONLINE HADR DATABASE ON THE OLD PRIMARY NODE. | |
Workaround | |
keiner bekannt / siehe Local-Fix | |
Weitere Daten | |
Datum - Problem gemeldet : Datum - Problem geschlossen : Datum - der letzten Änderung: | 25.01.2010 25.05.2010 25.05.2010 |
Problem behoben ab folgender Versionen (IBM BugInfos) | |
9.7.FP2 | |
Problem behoben lt. FixList in der Version | |
9.7.0.2 |