Understanding interface configurations for takeover

Types of takeover configurations

The following table lists the three types of interface configurations that you can enable in a cluster.

Interface type	Description
Shared	This type of interface supports both the local and partner nodes. It contains both the local node and partner node IP addresses. During takeover, it supports the identity of both nodes.
Dedicated	This type of interface only supports the node in which it is installed. It contains the local node IP address only and does not participate in network communication beyond local node support during takeover. It is paired with a standby interface.
Standby	This type of interface is on the local node, but only contains the IP address of the partner node. It is paired with a dedicated interface.

Note
Most cluster interfaces are configured as shared interfaces because they do not require an extra NIC. However, some Asynchronous Transfer Mode (ATM) configurations require interfaces to be configured as dedicated and standby.

Interface roles in normal and takeover modes

The following table shows the role of each interface type in normal and takeover mode.

Interface type	Normal mode	Takeover mode
Shared	Supports the identity of the local node	Supports the identity of both the local node and the failed node
Dedicated	Supports the identity of the local node	Supports the identity of the local node
Standby	Idle	Supports the identity of the failed node

Takeover configuration with shared interfaces

In the following configuration illustration, you use two NICs to provide the two interfaces.

If Node 1 fails, interface e0 on Node 1 stops functioning, but the secondary address on e0 on Node 2 handles the Node 1 network connection with the 230 network.

If Node 2 fails, e0 on Node 2 stops functioning, but e0 on Node 1 substitutes for the failed interface and handles the Node 2 network connection with the 230 network.

Takeover configuration with dedicated and standby interfaces

In the following configuration illustration, you use two NICs for each interface, one on each storage system. One NIC acts as a dedicated interface and the other acts as a standby interface.

If Node 1 fails, interface e0 on Node 1 stops functioning, but e0 on Node 2 substitutes for the failed interface and handles the Node 1 network connection with the 230 network.

If Node 2 fails, e1 on Node 2 stops functioning, but e1 on Node 1 substitutes for the failed interface and handles the Node 2 network connection with the 230 network.

Interface types and configurations

The following table lists the configurations supported by each type of interface in a cluster.

Interface	Shared	Dedicated	Standby	Partner parameter
Ethernet	X	X	X	IP address or interface name
Gigabit Ethernet	X	X	X	IP address or interface name
FDDI	X	X	X	IP address or interface name
Virtual interface	X	X	X	Virtual interface name
VLAN interface	X	X	X	IP address or interface name
ATM FORE/IP		X	X	IP address or FORE/IP interface name
ATM adapter for LANE	X¹	X	X	ATM adapter interface name

¹ATM adapters in a cluster can be shared for LANE provided that the two adapters do not have Emulated LANs in common.