keepalives

Keepalives are mechanisms used to verify that a network connection or path remains active between two endpoints. They help detect broken or idle connections, reclaim resources, and assist intermediaries such as routers, proxies, and NAT devices in maintaining state. Keepalives can operate at different layers, including the transport layer and application layer, and may take the form of periodic probes or policy-based messages.

In the transport layer, TCP keepalive is a low-level feature that sends periodic probes if no data has been exchanged for a configured idle period. If the peer fails to respond to a probe, after a number of attempts the connection is considered dead. Common parameters include the idle time before probes start, the interval between probes, and the maximum number of probes. Defaults vary by operating system and can be adjusted to balance timely detection against network overhead. TCP keepalives can prevent silent drops due to NAT or firewall timeouts but can increase unnecessary traffic and risk premature termination in congested networks.

Application-layer keepalives are often implemented as heartbeats or ping/pong messages to confirm liveness even when no user data is being sent. Examples include WebSocket ping/pong frames and custom heartbeat messages in chat or game servers. HTTP keep-alive, or persistent connections, allows reuse of a single TCP connection for multiple requests, reducing connection setup overhead; this is a separate mechanism from transport keepalives but complements connection longevity in HTTP traffic.

Considerations for using keepalives include network overhead, false positives, security implications (such as potential abuse or congestion during attacks), and the need for appropriate configuration on clients, servers, and intermediaries. Tools such as network monitors can help diagnose keepalive behavior.