In addition to normal startup, restart, and forced startup, the following startup procedures are also available:

• Startup procedure used to reinitialize a database
• Startup procedure used in the event of an error in the master directory RDAREA
• Startup procedure used in the event of an error in a data dictionary RDAREA

Note

You must exercise caution when you use the HA monitor's system switchover facility on an HP-UX or AIX edition of a HiRDB parallel server configuration. You must use the following procedure to release the HA monitor's monitoring status, and then you must execute a special startup procedure:

Procedure

1. Check the status of each unit (whether it is the running system or a standby system).^#
2. If the primary unit is in standby status, use the HA monitor's monswap command to switch the systems so that the primary unit becomes the running system. In other words, return the systems to their status when use of the system switchover facility began.
3. Enter the pdstop command to terminate HiRDB on the running system.
4. If a HiRDB is running as a server system having no HA monitor interface (in the monitor mode), use the monend command of the HA monitor to stop the HiRDB on the standby system.

#: If IP addresses are not inherited, the pdls -d ha command can be used to check the status.

Organization of this section

1.3.1 Startup procedure used to reinitialize a database (pdstart -i)

1.3.2 Startup procedure used in the event of an error in the master directory RDAREA (pdstart -r [-t])

1.3.3 Startup procedure used when the front-end server is in SUSPEND status due to an error in a data dictionary RDAREA (pdstart -a)