HP Serviceguard Enterprise Cluster Master Toolkit User Guide, June 2014

NOTE:
• For multiple physical and logical partition configuration of DB2, the number of ports
added in the services file must be sufficient for the number of partitions created in the
current node and the number of partitions created on the other nodes. This ensures that
enough ports are available for all partitions to startup on a single node, if all packages
managing different partitions are started on that node.
• For multiple physical and logical partition configuration of DB2, number of ports added
in services file must be sufficient for the number of partitions created in the current node
and the number of partitions created on the other nodes. This ensures that enough ports
are available for all partitions to startup on a single node, if all packages managing
different partitions are started on that node.
• If you own an instance, run the following:
[payroll_inst@node1/]> db2set DB2COMM=tcpip
[payroll_inst@node1/]> update database manager configuration using svcename db2c_payroll_inst
[payroll_inst@node1/]> db2stop
[payroll_inst@node1/]> db2start
• For multi-partitioned database using DB2 V9.7, a parameter called
DB2_PMODEL_SETTINGS must be set to SERIAL_RESTART:TRUE. This parameter along
with the respective Fix Pack installation, provides fix for Database Partitioning Feature
(DPF) issue reported by IC66748:DB2START RESTART NEEDS SERIAL FAILOVER CODE
RE-USE FROM V91 IN V97 case.
For example:
[payroll_inst@node1/]> db2set DB2_PMODEL_SETTINGS=SERIAL_RESTART:TRUE
11. Start the database manager and examine the database log for any errors.
[payroll_inst@node1 ~]> db2start
07/15/2008 18:10:32 1 0 SQL1063N DB2START processing was successful.
04/15/2008 18:10:32 0 0 SQL1063N DB2START processing was successful.
SQL1063N DB2START processing was successful.
12. Define the DBNAME and MNTPOINT variables. Create the database on the mount point defined,
connect to it locally, and terminate the connection.
As an instance owner, execute,
payroll_inst@node1 ~]> db2 create database $DBNAME on $MNTPOINT
payroll_inst@node1 ~]> db2 connect to $DBNAME (as instance owner)
Check the DB2 log for errors in
$HOME/sqllib/db2dump/db2diag.log
On other cluster nodes list the already created database:
payroll_inst@node1 ~]> db2 catalog database $DBNAME on $MNTPOINT \
authentication client
13. Terminate the connection and release the logical volume.
[payroll_inst@node1 ~]> db2 terminate (as instance owner)
[payroll_inst@node1 ~]> db2stop (as instance user)
[root@node1 /]> umount $MNTPOINT (as root user)
[root@node1 /]> vgchange -a n $VG (as root user)
14. Stop database manager and correct errors (if any).
[payroll_inst@node1 ~]> db2stop
Setting up the Application 77