HP Mainframe Connectivity Design Guide

1. The production LPARs write data to the production volumes, without disruption by the replication
process on the SDM LPARs.
2. The SDM reads the source storage system updates and queues them for writing on the target
storage system.
You can configure volumes from the source storage system or multiple storage systems into
nl
consistency groups.
3. When the SDM has all of the updates for a consistency group, it writes (commits) the data to
the target volumes.
4. When the commitment process is complete, the update process is repeated.
Figure 16 Sample XRC configuration
25292a
Production
LPARs
Production
volumes
(Source)
SDM
LPAR(s)
Replicated
volumes
(Target)
Local Remote
NOTE: The source storage system must be XRC certified. HP recommends that the target storage
system be XRC certified, but this is not a requirement.
XRC configuration support
HP supports XRC configurations with FICON SANs, including cascaded FICON directors.
The FICON SANs can be deployed between:
• Production LPARs and production volumes
• SDM LPARs and production volumes
• SDM LPARs and target volumes
The SDM mainframe can reside in the source data center or the remote data center. HP recommends
that the mainframe reside in the remote data center; this is a requirement if the XRC configuration
includes an FCIP gateway.
There are two types of XRC configurations: push and pull. HP recommends the pull configuration.
XRC pull configuration
In a pull configuration, the SDM mainframe maintains contact with target volumes, even if a link
fails. All in-progress updates are queued in the local storage system. After the link is reestablished,
the update process continues. Resynchronization of target volumes is not required in most cases;
however, with an extended link outage, resynchronization may be required. HP recommends that
you configure XRC with complete redundancy.
FICON long-distance extension applications 125