Difference between revisions of "Timeout"
From Computer History Wiki
m (Proper cat, fix error) |
m (Avoid redir) |
||
Line 1: | Line 1: | ||
In a [[packet switch]]ing [[communication network]], a '''timeout''' occurs when no [[acknowledgement]] for data or a [[packet]] sent by an entity to another is received from the other before a given time after it was originally sent - presumably because it was damaged or lost by the network. After the timeout happens, the data/packet will be [[retransmission|retransmitted]]. | In a [[packet switch]]ing [[communication network]], a '''timeout''' occurs when no [[acknowledgement]] for data or a [[packet]] sent by an entity to another is received from the other before a given time after it was originally sent - presumably because it was damaged or lost by the network. After the timeout happens, the data/packet will be [[retransmission|retransmitted]]. | ||
− | Since the delay is a real-time delay (since the entity at the other end is independent), such systems are a variety of [[real-time system]]. In a [[wide | + | Since the delay is a real-time delay (since the entity at the other end is independent), such systems are a variety of [[real-time system]]. In a [[wide area network]], calculating what the appropriate delay should be is a non-trivial problem, and it took well over a decade of work before the timers in the [[TCP/IP]] [[protocol suite]] worked well. |
{{semi-stub}} | {{semi-stub}} | ||
[[Category: Networking]] | [[Category: Networking]] |
Revision as of 20:17, 5 November 2018
In a packet switching communication network, a timeout occurs when no acknowledgement for data or a packet sent by an entity to another is received from the other before a given time after it was originally sent - presumably because it was damaged or lost by the network. After the timeout happens, the data/packet will be retransmitted.
Since the delay is a real-time delay (since the entity at the other end is independent), such systems are a variety of real-time system. In a wide area network, calculating what the appropriate delay should be is a non-trivial problem, and it took well over a decade of work before the timers in the TCP/IP protocol suite worked well.