Even if a disaster, such as an earthquake or fire, makes it difficult to physically recover the system you normally use, you can continue operations by switching to a secondary system that has been prepared at a remote location. The system environment that allows you to do this is called Real Time SAN Replication (RiSe). The site where the system you normally use is located is called the main site, and the remote site where the secondary system is located is called the remote site.

Data at the main site and the remote site is kept on a Hitachi disk array system, and if a change is made to the data at the main site, the TrueCopy or Universal Replicator feature of the Hitachi disk array system is used to import the changed data to the remote site (update copy).

The following figure provides an overview of Real Time SAN Replication.

Figure 1-1 Overview of Real Time SAN Replication

Explanation

• Normally, operations are performed using the HiRDB system at main site. When a file at the main site is updated, the updated content is copied to the remote site (update copy). Update copy keeps the data at the main site and the remote site synchronized.
• If a large-scale disaster, such as an earthquake or fire, occurs at the main site, making it impossible to quickly restore the system at the main site, you can continue operations by restarting HiRDB at the remote site.
  

  Reference note

  • Update copy is automatically performed by TrueCopy or Universal Replicator. TrueCopy and Universal Replicator copies data directly between linked Hitachi disk array systems without going through the hosts.
  • RAID Manager is an optional program product for Hitachi disk array systems and provides commands for controlling and operating TrueCopy and Universal Replicator.

Update copy targets the files listed below. Whenever these files are updated, the updated information is copied to the same files at the remote site.

• Database files (HiRDB files in RDAREAs)
• System log files
• Synchronization point dump files
• Status files

Update copy processing can be classified into synchronous copy and asynchronous copy. The table below shows the characteristics of synchronous copy and asynchronous copy.

Table 1-1 Characteristics of synchronous copy and asynchronous copy

Item	Synchronous copy	Asynchronous copy
Processing method	Updating at the main site is completed after updating at the remote site is completed (updating at the main site waits for updating at the remote site to be completed).	Updating at the main site is completed without waiting for updating at the remote site to be completed.
Data integrity between the main site and the remote site	Data at the main site always matches the data at the remote site.	Data loss might occur. Consequently, data at the main site might not match the data at the remote site.
Impact on performance#	Transaction processing performance is slowed. The amount of slowing is proportional to the distance between the sites.	There is no impact on performance.

#: Based on the theoretical performance of TrueCopy and Universal Replicator