Hitachi TrueCopy for IBM z/OS User and Reference Guide (T5211-96001, June 2007)

Preparing for TrueCopy for z/OS Operations 3-13
Hitachi Universal Storage Platform V TrueCopy for IBM z/OS User’s Guide
Performance and DR Considerations
Before starting TrueCopy for z/OS operations, you must consider the relative
importance of storage system I/O performance and disaster recovery
preparation. TCz operations can affect the I/O performance of the MCUs and
RCUs. TCz provides many options (RCU options, initial copy options, pair
options, group options, and async options) which allow you to control the
impact of TCz operations on storage system I/O performance. These options
can be set separately for each TCz pair, for each TCzA group, and for each
MCU to provide maximum flexibility. You can select options which minimize the
impact of TCz operations on storage system performance, or options which
maximize the effectiveness of TCz operations to ensure the best level of
backup data integrity. System-level factors (e.g., number of paths) can also
affect TCz operations and storage system performance (Table 5-7).
For TCz disaster recovery operations, you should make sure that the RCUs are
attached to a host processor to enable reporting of sense information and
transfer of ERC information. If the remote site is unattended, you should
attach the RCUs to a host processor at the main site, so that the system
administrator can monitor the operational condition of the RCUs.
Note: TrueCopy for z/OS and ShadowImage for z/OS can function together in
the same USP V to provide both internal and remote backup for your
important data. If you are planning to combine TrueCopy for z/OS and
ShadowImage for z/OS, please read the important configuration information in
Interoperability with Other Products and Functions
.
Identifying the Source and Target Volumes
To prepare for TCz pair operations, you need to identify the volumes (by LDEV
ID) for which you want to establish TCz remote copy operations. You should
identify not only the volumes which contain the important data to be backed
up (e.g., DB2
®
log files), but also the volumes which contain the catalog and
control datasets (i.e., master catalog, key user catalogs, system control
datasets). Copying these datasets to the remote site will enable faster disaster
recovery than maintaining a current version of these files at the remote site.
For large databases which span multiple volumes and even multiple USP Vs,
you should plan to establish a TCzA consistency group for each database, so
that the update sequence consistency of the database can be ensured at the
remote site.
You can start establishing the TCz pairs as soon as you have:
• Identified the volumes which will become the TCz M-VOLs and R-VOLs
• Ensured that all system and TCz requirements have been met
• Assessed TCz Requirements
• Completed hardware and software installation
• Configured the MCUs and RCUs for TCz operations