3.1 Basic concepts High-availability designs that do not use dynamic routing protocols (which are part of TCP/IP) achieve availability based on Local Area Networking (LAN) capabilities instead. (Such LAN capabilities are also often referred to as layer 2 functions because LANs constitute layers 1 and 2 of the seven-layer OSI networking model). The most fundamental of those capabilities is to be able to transfer an IP address from a failing (or failed) physical interface or stack to a different physical interface or stack. In an z/OS Communications Server environment, dynamic movement of an IP address (without the use of dynamic routing) is called ARP takeover and requires OSA-Express adapters (in QDIO mode). Note: The way ARP takeover and OSA-Express QDIO mode work together is introduced in 1.1.4, "z/OS network connectivity and dynamic routing" on page 5. An ARP takeover implementation example is provided in 3.5, "ARP takeover" on page 53. In addition, if you wish to move an IP address from one adapter to another within a system, you must also set up Static Virtual IP Addresses (VIPA). And, if you want to implement dynamic movement of an IP address from one TCP/IP stack to another (including a stack on a completely different system), you must set up Dynamic Virtual IP Addressing (DVIPA), which, in turn, requires Sysplex XCF communications. Note: Virtual IP Addressing is introduced, including implementation scenarios in Chapter 2, "Virtual IP Addressing" on page 13. ARP takeover uses LAN broadcast protocols; therefore, in order to use ARP takeover, all of the systems in your sysplex must be connected to the same LAN (or virtual LAN, VLAN), and,