HyperSwap configuration without using interswitch links

The most simple IBMHyperSwap configuration is built by attaching each node directly to the Fibre Channel switches in the local and the remote production site. This configuration is done without using interswitch links (ISLs).

HyperSwap configuration rules without using ISLs

The following rules must be observed when you configure a HyperSwap system without using ISLs:

For every storage system, create one zone that contains ports from every node and all storage system ports, unless the zoning guidelines for that storage system state differently. However, do not connect a storage system in one site directly to a switch fabric in the other site. Instead, connect each storage system only to switched fabrics in the local site.
Each node must have two direct Fibre Channel connections to one or more SAN fabrics at both locations that contain nodes.
You can include powered components between the system and the switches in a HyperSwap system configuration. For example, you can use powered dense wavelength division multiplexing (DWDM) Fibre Channel extenders.
Be prepared to provide and replace longwave SFP transceivers in case they fail.
Some service actions require the ability to do actions on the front panel or through the technician port of all nodes in a system within a short-time window. If you use HyperSwap systems, you are required to assist the support engineer and provide communication technology to coordinate these actions between the sites.

Configure the HyperSwap system that does not include interswitch links (ISLs) according to the following rules:

The minimal SAN configuration consists of one Fibre Channel switch per production site as two separate fabrics. For highest reliability, use two switches per production site. Single fabric configurations are not supported for HyperSwap I/O group systems.
Note: FCIP connectivity is not supported between nodes when a HyperSwap system is configured without using ISLs. FCIP (or any Fibre Channel extender or SAN router that is supported for intersystem communications) might be used in the connection to the quorum disk at the third site.
FCIP connectivity is supported (or any Fibre Channel extender or SAN router that is supported for intersystem communications) both between nodes at two different sites and between nodes and the quorum disk at the third site when a HyperSwap system is configured with ISLs.
As with every clustered system, you can use ISLs for host-to-node (with up to three hops) or for node-to-storage (at most one hop) access. However, configure the SAN zones so that ISLs are not used in paths between nodes.
In a system with four ports per node, attach half of the ports of each node to the Fibre Channel switches in the production site where the node is located.
In the same system with four ports per node, attach the remaining half of the ports of each node to the Fibre Channel switches in the other (remote) production site.
Connect each storage system at the production sites to Fibre Channel switches in the site where the storage system is located.
Connect the storage system with the active quorum disks to Fibre Channel switches in both production sites.
To avoid fabric topology changes in case of IP errors, it is a good practice to configure FCIP links so that they do not carry ISLs.
Note: This guideline applies only if FCIP is used to connect to a switch at the third site where the quorum disk is located.

Restriction: Do not connect a storage system in one site directly to a switch fabric in the other site. Connectivity between system ports at one site to storage at the other site that uses ISLs to travel between the locations are not affected by this restriction.

It is strictly required that the links from both production sites to the quorum site are independent and do not share any long-distance equipment.