A customer requires a high-availability SAN for disk access. The disk array and server may each have multiple ports. How many Fabric switch ports are…

To achieve full redundancy and eliminate any single point of failure (SPOF), a dual-fabric SAN topology is the standard design. This architecture uses two independent switches, creating two separate data paths. The server must connect to both fabrics, requiring a minimum of two host bus adapters (HBAs) or ports. Likewise, the disk array must connect to both fabrics, requiring at least two controller ports. Each switch will therefore have one connection to the server and one to the disk array. This configuration uses a total of four switch ports (2 ports per switch x 2 switches), ensuring that the failure of any single component—HBA, cable, switch, or array port—does not sever access to the storage.

A. 3: This number is insufficient to create two fully independent and redundant data paths required for a high-availability, dual-fabric design.

C. 5: While functional, this is more than the minimum number of ports necessary to satisfy the requirement for full redundancy with one server and one array.

D. 6: This is also more than the minimum number of ports required. Additional ports would be for increased throughput or connecting more devices, which is not the focus here.

---

1. IBM Redbooks Publication, "Introduction to Storage Area Networks," SG24-5470-06.

Reference: Chapter 3, "Fibre Channel," Section "Fibre Channel topologies," page 70, Figure 3-18 "Dual-fabric SAN".

Content: This official IBM guide illustrates a dual-fabric SAN designed for high availability. The diagram shows hosts and storage subsystems connecting to two separate fabrics (Fabric A and Fabric B). For a single host and a single storage system, this topology requires one connection from each device to each of the two switches, totaling four switch ports.

2. Cisco, "Cisco MDS 9000 Series High Availability Configuration Guide, Release 9.3(x)."

Reference: Chapter: "High Availability," Section: "Redundancy."

Content: This official vendor documentation explains that high availability in a SAN is achieved through redundancy at all levels, including "redundant fabrics." It states, "A single switch is a single point of failure. A second switch and redundant paths from the host to the storage through the second switch provide the required redundancy." This principle directly leads to the requirement of two switches and four total ports for the scenario.

3. University of Pittsburgh, Course INFSCI 2160: Data Warehousing, Lecture 10: "Storage Area Networks."

Reference: Slide 17, "Redundant SAN Fabric."

Content: The courseware diagram explicitly shows a redundant SAN fabric with two switches. A host server is shown with two HBAs, each connecting to a different switch. The storage array is also shown with two controllers, each connecting to a different switch. This visual representation confirms the need for four switch ports to connect one server and one array in a fully redundant manner.