Nonstop Database, HiRDB Version 9 System Operation Guide
This subsection provides examples of system configurations when the system switchover facility is used.
A server in a HiRDB single server configuration performs system switchovers at the system level. The following subsections provide examples of system configurations for a HiRDB single server configuration.
A configuration in which there is a one-to-one correspondence between the running system and the standby system is called a 1-to-1 switchover configuration. This configuration is appropriate in cases where it is important to guarantee response time when a system switchover becomes necessary. The drawback of this configuration is that you cannot utilize the resources of the server machine in the standby system (you do not have access to the resources of one of your two server machines). The following figure shows a system configuration example of a HiRDB single server configuration (1-to-1 switchover configuration).
Figure 26-47 System configuration example of a HiRDB single server configuration (1-to1 switchover configuration)
A system configuration in which the running system has a system on the same server machine that acts as a mutual standby system (standby system for another HiRDB single server configuration) is called a mutual system switchover configuration. You can apply this configuration when multiple HiRDB single server configurations are being used. This configuration is appropriate for making efficient use of server machine resources. However, response time slows when a system switchover occurs.
The figure below shows a system configuration example of HiRDB single server configurations (mutual system switchover). In this example, a mutual system switchover configuration is applied to two HiRDB single server configurations (HiRDB single server configuration A and HiRDB single server configuration B).
Figure 26-48 System configuration example of HiRDB single server configurations (mutual system switchover)
A system configuration in which a single running system has multiple standby systems is called a multi-standby configuration. Use this configuration when you want to guard against errors that might occur in a standby system while the running system is recovering from an error (multi-point error). The standby systems are assigned different priority levels, and operation is switched to the highest-priority standby system when an error occurs in the running system. Note that the multi-standby configuration can be applied only to a system configuration in which IP addresses are inherited.
The following figure shows an example of a multi-standby configuration for a HiRDB single server configuration.
Figure 26-49 Example of a multi-standby configuration for a HiRDB single server configuration
The figure below shows an example of sharing a utility special unit among multiple HiRDB single server configurations. Note that system switchovers cannot be performed on utility special units.
Figure 26-50 Sharing a utility special unit among multiple HiRDB single server configurations
The following figure shows an example of setting up a 1:1 correspondence between HiRDB single server configurations and utility special units.
Figure 26-51 Setting up a 1:1 correspondence between HiRDB single server configurations and utility special units
The following figure shows an example of setting up an m:n correspondence between HiRDB Single Servers and utility special units.
Figure 26-52 Setting up a m:n correspondence between HiRDB single server configurations and utility special units
A server in a HiRDB parallel server configuration performs system switchovers at the unit level. The following subsections provide examples of system configurations for a HiRDB parallel server configuration.
A configuration in which there is a one-to-one correspondence between the running system units and the standby system units is called a 1-to-1 switchover configuration. This configuration is appropriate in cases where it is important to guarantee response time when a system switchover becomes necessary. The drawback of this configuration is that you cannot utilize the resources of the server machine in the standby system (you do not have access to the resources of one of your two server machines). The following figure shows a system configuration example of a HiRDB single server configuration (1-to-1 switchover configuration).
Figure 26-53 System configuration example of a HiRDB parallel server configuration (1-to-1 switchover configuration)
A system configuration in which the running system has a system on the same server machine that acts as its mutual standby system (standby system for another unit) is called a mutual system switchover configuration. In a HiRDB parallel server configuration, systems can be switched over in units. Therefore, whereas a mutual system switchover configuration cannot be applied to a single HiRDB single server configuration, it can be applied to a single HiRDB parallel server configuration. When a mutual system switchover configuration is applied to a HiRDB parallel server configuration, the running system unit and the standby system unit (standby system unit for another unit) can both be located on the same server machine. This configuration is appropriate for making efficient use of server machine resources. However, response time slows when a system switchover occurs.
The following figure shows a system configuration example of a HiRDB parallel server configuration (mutual system switchover).
Figure 26-54 System configuration example of a HiRDB parallel server configuration (mutual system switchover)
For examples of the HiRDB system definitions for a mutual system switchover configuration, see HiRDB parallel server configuration: when the standby system switchover facility is used in the manual HiRDB Version 9 System Definition.
The following figures show correct and incorrect examples of host name setup.
Figure 26-55 Example of correct host name setup
The following is an example of correct pdunit operand specifications:
pdunit -x hostA -u UNT1 ... -c hostAA pdunit -x hostB -u UNT2 ... -c hostBB |
Furthermore, you must specify the host names corresponding to different IP addresses for hostA, hostAA, hostB, and hostBB.
Figure 26-56 Example of incorrect host name setup
The following is an example of incorrect pdunit operand specifications:
pdunit -x hostA -u UNT1 ... -c hostB pdunit -x hostB -u UNT2 ... -c hostA |
A system configuration in which a single running system unit has multiple standby system units is called a multi-standby configuration. Use this configuration when you want to guard against errors that might occur in a standby system while the running system is recovering from an error (multi-point error). The standby systems are assigned different priority levels, and operation is switched to the highest-priority standby system when an error occurs in the running system. Note that the multi-standby configuration can be applied only to a system configuration in which IP addresses are inherited. The following figure shows an example of a multi-standby configuration for a HiRDB parallel server configuration.
Figure 26-57 Example of a multi-standby configuration for a HiRDB parallel server configuration
All Rights Reserved. Copyright (C) 2011, 2015, Hitachi, Ltd.