Goodput

<h2 id="example">Example</h2>
<p>Over Ethernet, files are broken down into individual chunks for transmission. These chunks are no larger than the <a href="/facts/Maximum_transmission_unit/l2FdKoLa">maximum transmission unit</a> of IP over Ethernet, or 1500 <a href="/facts/Byte/BNHUp9Qq">bytes</a>. Each packet requires 20 bytes of <a href="/facts/Internet_Protocol/Qq6VeEal">IPv4</a> header information and 20 bytes of <a href="/facts/Transmission_Control_Protocol/WW4sPrt9">TCP</a> header information, leaving 1460 bytes per packet for file data (<a href="/facts/Linux/9FvCKSTq">Linux</a> and <a href="/facts/MacOS/Y93vVpTm">macOS</a><a class="footnote-ref" id="fnref:1" href="#fn:1"><sup>1</sup></a> are further limited to 1448 bytes as they also carry a 12-byte time stamp). The data is transmitted over Ethernet in a frame, which imposes a 26 byte overhead per packet. Given these overheads, the maximum <i>goodput</i> is 1460/1526 × 100 <a href="/facts/Mbit/s/Nv9VoT7T">Mbit/s</a> which is 95.67 megabits per second or 11.959 <a href="/facts/Megabyte_per_second/Nv9VoT7T">megabytes per second</a>.
</p><p>Note that this example doesn't consider additional Ethernet overhead, such as the <a href="/facts/Interframe_gap/7BU21IKz">interframe gap</a> (a minimum of 96 bit times), or collisions (which have a variable impact, depending on the network load). TCP adds the overhead of acknowledgements (which along with the <a href="/facts/Round-trip_delay_time/jYgNMK98">round-trip delay time</a> and the TCP window size in effect rate-limit each individual TCP connection, see <a href="/facts/Bandwidth-delay_product/72txR7o3">bandwidth-delay product</a>). This example does not consider the overhead of the HTTP protocol itself, which becomes relevant when transferring small files.
</p>
<h2 id="data-delivery-time">Data delivery time</h2>
<p>The goodput is a ratio between delivered amount of information, and the total delivery time. This delivery time includes:
</p>
<ul><li>Inter-packet time gaps caused by packet generation processing time (a source that does not use the full network capacity), or by protocol timing (for example <a href="/facts/Collision_avoidance_(networking)/hVMvKYeF">collision avoidance</a>)</li>
<li>Data and overhead <a href="/facts/Transmission_delay/kvtQ8glB">transmission delay</a> (amount of data divided by <a href="/facts/Bit_rate/ppUNMmVd">bit rate</a>)</li>
<li><a href="/facts/Propagation_delay/tUcrAwwe">Propagation delay</a> (distance divided by wave propagation speed)</li>
<li>Packet <a href="/facts/Queuing_delay/Kbr94SPe">queuing delay</a></li>
<li><a href="/facts/Network_address_translation/nFweFNfN">NAT</a> translation delay</li>
<li>Intermediate node <a href="/facts/Store_and_forward_delay/kvtQ8glB">store-and-forward processing delay</a></li>
<li>Packet retransmission time (in case of deleted packets in congested routers, or detected bit errors)</li>
<li>Delayed acknowledge due to flow control, <a href="/facts/Congestion_avoidance/9ohkRMso">congestion avoidance</a> and <a href="/facts/Processing_delay/nAUYRSwV">processing delay</a></li></ul>
<h2 id="see-also">See also</h2>
<ul><li><a href="/facts/Measuring_network_throughput/l90avxEk">Measuring network throughput</a></li>
<li>Link and system <a href="/facts/Spectral_efficiency/GaePAvv9">spectral efficiency</a></li></ul>

<h2 id="sources">Sources</h2>
<ul><li><a href="http://mailman.isi.edu/pipermail/ns-users/2001-June/015799.html">Goodput calculation</a> <a href="https://web.archive.org/web/20100818222126/http://mailman.isi.edu/pipermail/ns-users/2001-June/015799.html">Archived</a> 2010-08-18 at the <a href="/facts/Wayback_Machine/nmQ3a6JC">Wayback Machine</a></li>
<li><a href="https://arxiv.org/abs/cs.IT/0604082">Energy-Efficient Power and Rate Control with QoS Constraints: A Game-Theoretic Approach</a></li>
<li><a href="http://tools.ietf.org/html/rfc2647">RFC 2647 — Benchmarking Terminology for Firewall Performance</a></li></ul>

<h2 id="references">References</h2>

<ol>
<li id="fn:1"><p>Stuart Cheshire. "TCP Performance problems caused by interaction between Nagle's Algorithm and Delayed ACK". Retrieved 2010-01-13. <a href="http://www.stuartcheshire.org/papers/NagleDelayedAck/index.html" target="_blank">http://www.stuartcheshire.org/papers/NagleDelayedAck/index.html</a> <a href="#fnref:1" class="footnote-back-ref">↩</a></p></li>
</ol>

Goodput open-in-new

Goodput