Partnerships
Partnerships are used to connect systems together to enable migration, data replication, and high-availability solutions.
A system can have partnerships with up to three remote systems. The connectivity for each partnership can be either Fibre Channel or IP. Systems also become indirectly associated with each other through partnerships. If two systems each have a partnership with a third system, those two systems are indirectly associated. A maximum of four systems can be directly or indirectly associated with each other.
A partnership configuration requires actions on both systems involved. This ensures that there is authority to access each system and share data between them.
- By using Fibre Channel connectivity.
- By using Partnerships using IP Connectivity.
Policy-based Replication
To use partnerships for policy-based replication (asynchronous or high-availability), IP connectivity is required between management IP addresses of partnered systems.
Replication management requires access to the REST API on the remote system. Replication setup using the management GUI requires access to the management GUI of the remote system. Ensure that firewalls between the systems allow inbound traffic to port 7443 and 443 on the system management IP address.
The management traffic uses authentication certificates to prevent unauthorized access and ensure secure communications between the systems. Therefore, ensure that valid authentication certificates are installed on both the systems. For more information, see Configuring system certificates.
Background copy management
Certain types of replication differentiate between foreground host writes and background synchronisation traffic. The background copy rate is specified as a percentage of the partnership link bandwidth that is available to background synchronisation activities. Policy-based replication treats all traffic as background work so the background copy rate should be set to 100% if the system is using only policy-based replication.
HyperSwap, Metro Mirror, and Global Mirror (if supported by your system) use the background copy rate to control synchronisation traffic. Multi-cycling Global Mirror (Global Mirror with Change Volumes) uses only background copy therefore to achieve the best possible recovery point the background copy rate should be set to 100%. If you are using Metro Mirror, HyperSwap or non-cycling Global Mirror on your system, a lower value should be used to ensure that there is sufficient bandwidth to replicate host writes.
Replication between IBM Storage Virtualize systems
- A SAN Volume Controller system is always in the replication layer.
- A FlashSystem is in the replication layer by default, but the system can be configured to be in the storage layer instead.