Cisco MDS 9000 Family NX-OS Interfaces Configuration Guide

FlexAttach Virtual pWWN

FlexAttach virtual pWWN feature facilitates server and configuration management. In a SAN environment, the server installation or replacement, requires interaction and coordination among the SAN and server administrators. For coordination, it is important that the SAN configuration does not change when a new server is installed, or when an existing server is replaced. FlexAttach virtual pWWN minimizes the interaction between the server administrator and the SAN administrator by abstracting the real pWWN using virtual pWWNs.

When FlexAttach virtual pWWN is enabled on an interface, a virtual pWWN is assigned to the server interface. The real pWWN is replaced by a virtual pWWN, which is used for a SAN configuration such as zoning.

Server administrators can benefit from FlexAttach in the following scenarios:

• Pre-configure—Pre-configure SAN for new servers that are not available physically yet. For example, they may be on order. FlexAttach can be enabled on the ports designated for the new servers and use the virtual WWNs assigned for configuring SAN. The new servers are then plugged into the fabric without any change needed in the SAN.
• Replacement to the same port—A failed server can be replaced onto the same port without changing the SAN. The new server gets a same pWWN as the failed server because the virtual pWWN is assigned to the port.
• Replacement to (spare)—A spare server, which is on the same NPV device or a different NPV device) can be brought online without changes to the SAN. This action is achieved by moving the virtual port WWN from the current server port to the spare port.
• Server Mobility—A server can be moved to another port on the same NPV device or another NPV device without changing the SAN. This is accomplished by moving the virtual pWWN to the new port. No change is needed if FlexAttach was configured using the physical port WWN of the server to the virtual port WWN mapping.

Difference Between San Device Virtualization and FlexAttach Port Virtualization

Table 8-1 describes the difference between SAN device virtualization (SDV) and FlexAttach port virtualization.

Table 8-1 Difference Between SDV and FlexAttach Virtualization

FlexAttach Virtual pWWN CFS Distribution

The FlexAttach virtual pWWN configuration is distributed for CFS through IPv4, and is enabled by default. The FlexAttach virtual pWWN distribution, by default, is on CFS region 201. The CFS region 201 links only to the NPV-enabled switches. Other CFS features such as syslog is on region 0. Region 0 will be linked through IPv4 for all NPV switches on the same physical fabric. If CFS has an option to link through IPv4 or ISL, then CFS will select the ISL path.

Note NPV switches do not have ISL (E or TE ports) and are linked through IPv4.

Security Settings for FlexAttach Virtual pWWN

Security settings for the FlexAttach virtual pWWN feature are done by port security at the NPV core. Node WWN of the end device is used to provide physical security.

For more details on enabling port security, refer to the Cisco MDS 9000 Series NX-OS Security Configuration Guide.

Automatically Assigning FlexAttach Virtual pWWN

Automatic assignment of virtual pWWN can be configured on an NPV switch globally, per VSAN, or per port. When assigned automatically, a virtual WWN is generated from the device local switch WWN.

Prerequisites

The port must be in a shut state when the virtual pWWN is enabled.

Detailed Steps

To assign a virtual pWWN automatically, perform this task:

Troubleshooting Tips

• When the interface-list value is not included in the command, virtual pWWN is assigned globally.
• All the interfaces mentioned in the interface-list value must be in a shut state.

Manually Assigning FlexAttach Virtual pWWN

You can manually assign a WWN to the interface, without generating it through the switch. Several checks are done by the NPV core to ensure the uniqueness of virtual pWWNs in the switch. When duplicate virtual pWWNs are configured, the subsequent logins are rejected by the NPV core switch.

Prerequisites

• Some ports may be in automode, some in manual mode, and the virtual pWWNs need not be assigned.
• The port must be in a shut state when a virtual pWWN is assigned.

Restrictions

• The interface mentioned in the interface value must be in a shut state.

Detailed Steps

To assign virtual pWWN manually, perform this task:

Mapping pWWN to Virtual pWWN

You can configure virtual pWWNs through real pWWNs. This process is required for NPIV hosts containing multiple pWWNs, of which only FLOGI is mapped to the virtual pWWN. Subsequent FDSIDs will have different mappings.

Several checks are done by the NPV core to ensure the uniqueness of virtual pWWNs in the switch across the NPV switches. When duplicate virtual pWWNs are configured, the subsequent logins are rejected by the NPV core switch.

Prerequisites

• The interface must be in a shut state and the specified virtual pWWN should not be logged in.

Restrictions

• The specified virtual pWWN and the real pWWN must not be logged in.

Detailed Steps

To map pWWN to virtual pWWN, perform this task:

Verifying the End Device

To verify that the end device is logged with the correct virtual WWNs, use the show fcns database command on the NPV core. (See Example 8-2.)

Example 8-2 Verifying the End Device

Total number of entries = 4