rebind

Rebind is a term used in multiple technical contexts to describe the act of binding something again or binding to a different target.

In programming, rebind most commonly appears in the context of allocator design in C++. The allocator concept in older C++ standards exposed a nested member template called rebind, which provides an allocator for another type from an existing allocator. For example, std::allocator<T>::rebind<U>::other yields an allocator<U> that shares the same underlying allocation strategy. With the advent of allocator_traits in modern C++, explicit rebind usage has diminished, but the concept remains in legacy code and discussions of allocator compatibility.

In computer networking, rebinding can refer to the process of binding a socket to a (potentially different) local address or port, sometimes by closing and reopening the socket or by altering binding options. The operation is distinct from the initial bind, and its feasibility and behavior depend on the operating system and socket options such as SO_REUSEADDR or SO_REUSEPORT. Rebinding issues can arise when addresses are in use, when interfaces change, or when a program needs to switch the local endpoint it uses for outgoing or incoming connections.

In security and web contexts, DNS rebinding describes a class of attacks where an attacker manipulates DNS responses to cause a victim’s browser to make requests to a private internal IP address. This can bypass certain network protections and violate same-origin policies. Mitigations include enforcing strict same-origin policies, limiting DNS caching behavior, and implementing network-level or application-level boundaries to prevent unauthorized access to internal services.

Beyond these domains, rebind appears as a general term in discussions of reattaching, reassigning, or reestablishing bindings in various systems and APIs.