Stayin' Dead

Understanding TCP keepalive and its legitimate use cases

This is a hard networking fundamentals question that tests whether you understand the actual purpose and constraints of TCP keepalive mechanisms. It's the kind of question that separates candidates who have worked with low-level socket programming from those who only know networking in theory.

TCP keepalive exists to detect dead connections and refresh router state, but it has specific limitations and is not a universal solution for every network problem. This question requires you to distinguish between the genuine reasons keepalive was designed for and the misconceptions candidates often have about what it can accomplish. You'll need to reason about protocol semantics, the OSI stack, and practical constraints on timing and overhead.

• Detecting unresponsive peers and dropped connections
• Refreshing NAT/firewall session tables
• The difference between application-level and transport-level heartbeats
• Trade-offs between keepalive overhead and detection latency