pinning

Pinning is the act of fixing or securing a thing in a chosen position or state, often using a pin or analogous mechanism. In computing and data systems, pinning describes methods to prevent movement, loss, or eviction, and to designate priority or persistence for certain objects.

In memory management, page or memory pinning is used to prevent specific pages from being moved or swapped out to disk, ensuring stable access to memory regions, typically for hardware I/O or high-performance tasks. In processor and virtualization contexts, CPU pinning (also called CPU affinity) binds a process or thread to one or more specific CPU cores to improve cache locality and performance. Virtualized environments may also pin virtual CPUs to host cores to reduce scheduling overhead and improve predictability.

Security pinning refers to tying a domain to expected cryptographic keys or certificates. TLS or public key pinning aims to prevent man-in-the-middle attacks by detecting unexpected keys for a host. While once advocated in some browsers and apps, public key pinning has proven risky in practice and is largely deprecated in browsers; alternative approaches emphasize robust certificate management and certificate transparency.

In distributed and content-addressed systems, pinning is used to preserve data copies locally. For example, in IPFS, pinning marks objects to be retained on a node, preventing automatic garbage collection from removing them. Pin sets can be managed to ensure long-term availability by multiple nodes or pinning services.

On social and collaboration platforms, pinning refers to placing a post, thread, or item at the top of a list or board for prominence. Examples include pinning a message in a forum, a post in a profile timeline, or a bookmark-like pin on content platforms such as Pinterest.

Overall, pinning denotes fixing, prioritizing, or preserving elements within physical, software, or social systems, with specific implementation and risks varying by context.