(a) One-to-one system switchover configuration

This configuration provides a one-to-one correspondence between running systems and standby systems. Use this configuration when you wish to guarantee response times, even when a system has been switched over. However, you cannot use the resources on the standby server machines (one set of server machine resources cannot be used for every two server machines). Figure 8-4 illustrates a one-to-one system switchover configuration.

Figure 8-4 One-to-one system switchover configuration

(b) Two-to-one system switchover configuration

This configuration provides a two-to-one correspondence between running systems and standby systems. The secondary system is configured as a multi-HiRDB system. Use this configuration with operations for which you wish to guarantee response times, even when a system has been switched over (response times deteriorate, though, if the two running systems have both been switched over). However, you cannot use the resources on the standby server machines (one set of server machine resources cannot be used for every three server machines). Figure 8-5 illustrates a two-to-one system switchover configuration.

Figure 8-5 Two-to-one system switchover configuration

(c) Mutual system switchover configuration

This configuration provides an alternating standby system on each server machine while, at the same time, the server machine is operating as a running system. Every server machine is configured as a multi-HiRDB system consisting of a HiRDB running system and a HiRDB standby system. Use this configuration when you wish to utilize server machine resources most efficiently. However, response times deteriorate when a system has been switched over. Figure 8-6 illustrates a mutual system switchover configuration.

Figure 8-6 Mutual system switchover configuration

This subsection provides examples of typical system configurations when the standby-less system switchover (1:1) facility is used.

(a) Mutual alternating configuration

This configuration example employs reciprocal alternate BESs on two back-end servers using standby-less system switchover (1:1). Figure 8-7 illustrates a system configuration example of such a mutual alternating configuration.

Figure 8-7 System configuration example of a mutual alternating configuration

Explanation

• BES1 is the alternate BES for BES2. If a failure occurs on BES2, the alternate portion for BES2 takes over the BES2 processing.
• BES2 is the alternate BES for BES1. If a failure occurs on BES1, the alternate portion for BES1 takes over the BES1 processing.

(b) One-way alternating configuration

This configuration example employs an alternate BES on one back-end server only using standby-less system switchover (1:1). Figure 8-8 illustrates a system configuration example of a one-way alternating configuration (2-node configuration).

Figure 8-8 System configuration example of a one-way alternating configuration (2-node configuration)

Explanation

BES2 is the alternate BES for BES1. If a failure occurs on BES1, the alternate portion for BES1 takes over the BES1 processing. If a failure occurs on BES2, BES1 does not take over its processing.

Figure 8-9 illustrates a system configuration example of the standby-less system switchover (effects distributed) facility. When a failure occurs in a regular unit, processing directed to the back-end servers of the failed primary system is distributed to and executed on multiple active server machines at their back-end servers.

Figure 8-9 System configuration example of the standby-less system switchover (effects distributed) facility

Explanation

1. If a failure occurs in unit 1, unit 2 executes the processing as a guest BES for BES1 and unit 3 executes the processing as a guest BES for BES2.
2. If a failure occurs in unit 2 while unit 1 is still down, unit 3 executes the processing as guest BESs for BES1, BES2, BES3, and BES4.

Figure 8-10 shows a configuration example of a mixed standby-less (1:1) and standby type setup.

Figure 8-10 System configuration example of a mixed standby-less (1:1) and standby type setup

Explanation

• The units containing a MGR (system manager), FES (front-end server), and DS (dictionary server) employ the standby system switchover facility in a mutual system switchover configuration.
• The units containing a BES (back-end server) employ the standby-less system switchover (1:1) facility in a mutual alternating configuration.
• HiRDB Advanced High Availability is needed on all server machines. HiRDB Advanced High Availability is also required on server machines for which neither the standby-less system switchover facility nor the standby system switchover facility is employed.