The standby-less system switchover facility implements a system switchover method that enables HiRDB to continue providing services when a failure occurs by transferring processing to another unit. Because the standby-less system switchover facility does not require that standby resources be placed in reserve, it is significantly more economical than the previously available system switchover facility (the standby system switchover facility). With the previous system switchover facility, you had to set aside a server, CPU, and memory resources to be available for transfer of operations in the event of a failure. With the standby-less system switchover facility, there is no need to provide these redundant resources. You simply register another server as the alternate, and that server's unit takes over processing if a failure occurs. Although the processing performance of the unit that has taken over might be degraded, the overall cost of the system is reduced because you can make more effective use of resources.

The standby-less system switchover facility includes the standby-less system switchover (1:1) facility and the standby-less system switchover (effects distributed) facility. For details about the standby-less system switchover facility, see 8.1 System switchover facility.

Previously, changing HiRDB system definitions meant having to stop HiRDB; however, by using the system reconfiguration command (pdchgconf command), you can change HiRDB system definitions without having to stop HiRDB. This allows you to perform system-related operations while HiRDB is running, such as changing the configuration of units or servers and adding system files. For details about the system reconfiguration command, see 6.5.4 System reconfiguration command (pdchgconf command).

Previously, adding, changing, or removing global buffers required you to change the pdbuffer operand in the HiRDB system definitions, which meant you had to stop HiRDB. Using the pdbufmod command, however, allows you to add, change, and remove global buffers while HiRDB is running. This is called dynamic updating of global buffers. For details about dynamic updating of global buffers, see 6.8.1(3) Dynamic updating of global buffers.

Table matrix partitioning is a function that performs key range partitioning by using multiple columns as keys. The use of multiple columns as partitioning keys improves the performance of SQL code that is executed in parallel, and the use of multiple keys for searches accelerates processing by narrowing the searched range. In addition, partitioning database storage areas (RDAREAs) into smaller segments also reduces the time it takes to reorganize the database, to make backups, and to recover the database. For details about table matrix partitioning, see 3.3.9 Table matrix partitioning.

Using ALTER TABLE, you can change the partitioning storage conditions of tables that have been row-partitioned by key range partitioning.^# Changing a table's partitioning storage conditions enables you to reuse RDAREAs that contain outdated data and reduces work time. For details about changing partitioning storage conditions, see 3.3.10 Changing the partitioning storage conditions of a table.

#

ALTER TABLE can be used to change a table's partitioning storage condition when one of the following partitioning methods was used:

• Boundary value specification
• Storage condition specification (if the equal sign (=) was used as the comparison operator for the storage condition)