rfc9318v3.txt   rfc9318.txt 
skipping to change at line 715 skipping to change at line 715
End-to-end latency is the time that a particular packet takes to End-to-end latency is the time that a particular packet takes to
traverse the network path from the user to their destination and traverse the network path from the user to their destination and
back. The end-to-end latency comprises several components: back. The end-to-end latency comprises several components:
1. The propagation delay, which reflects the path distance and the 1. The propagation delay, which reflects the path distance and the
individual link technologies (e.g., fiber vs. satellite). The individual link technologies (e.g., fiber vs. satellite). The
propagation doesn't depend on the utilization of the network, to propagation doesn't depend on the utilization of the network, to
the extent that the network path remains constant. the extent that the network path remains constant.
2. The buffering delay, which reflects the time segments spend in 2. The buffering delay, which reflects the time segments spent in
the memory of the network equipment that connect the individual the memory of the network equipment that connect the individual
network links, as well as in the memory of the transmitting network links, as well as in the memory of the transmitting
endpoint. The buffering delay depends on the network endpoint. The buffering delay depends on the network
utilization, as well as on the algorithms that govern the queued utilization, as well as on the algorithms that govern the queued
segments. segments.
3. The transport protocol delays, which reflect the time spent in 3. The transport protocol delays, which reflect the time spent in
retransmission and reassembly, as well as the time spent when the retransmission and reassembly, as well as the time spent when the
transport is "head-of-line blocked". transport is "head-of-line blocked".
skipping to change at line 789 skipping to change at line 789
until a bottleneck capacity is reached. From these measurements, a until a bottleneck capacity is reached. From these measurements, a
round-trip latency is measured and reported to the end user. The round-trip latency is measured and reported to the end user. The
authors chose to report results with the RPM metric. The methodology authors chose to report results with the RPM metric. The methodology
had been implemented in Apple's macOS Monterey. had been implemented in Apple's macOS Monterey.
[Mathis2021] applied the RPM metric to the results of more than 4 [Mathis2021] applied the RPM metric to the results of more than 4
billion download tests that M-Lab performed from 2010-2021. During billion download tests that M-Lab performed from 2010-2021. During
this time frame, the M-Lab measurement platform underwent several this time frame, the M-Lab measurement platform underwent several
upgrades that allowed the research team to compare the effect of upgrades that allowed the research team to compare the effect of
different TCP congestion control algorithms (CCAs) on the measured different TCP congestion control algorithms (CCAs) on the measured
end-to-end latency. The study showed that the use of Cubic CCA leads end-to-end latency. The study showed that the use of cubic CCA leads
to increased working latency, which is attributed to its use of to increased working latency, which is attributed to its use of
larger queues. larger queues.
[Schlinker2019] presented a large-scale study that aimed to establish [Schlinker2019] presented a large-scale study that aimed to establish
a correlation between goodput and QoE on a large social network. The a correlation between goodput and QoE on a large social network. The
authors performed the measurements at multiple data centers from authors performed the measurements at multiple data centers from
which video segments of set sizes were streamed to a large number of which video segments of set sizes were streamed to a large number of
end users. The authors used the goodput and throughput metrics to end users. The authors used the goodput and throughput metrics to
determine whether particular paths were congested. determine whether particular paths were congested.
skipping to change at line 1026 skipping to change at line 1026
4.3.5. Efficient Collaboration between Hardware and Transport Protocols 4.3.5. Efficient Collaboration between Hardware and Transport Protocols
With the advent of the low latency, low loss, and scalable throughput With the advent of the low latency, low loss, and scalable throughput
(L4S) congestion notification and control, there is an even higher (L4S) congestion notification and control, there is an even higher
need for the transport protocols and the underlying hardware to work need for the transport protocols and the underlying hardware to work
in unison. in unison.
At the time of the workshop, the typical home router uses a single At the time of the workshop, the typical home router uses a single
FIFO queue that is large enough to allow amortizing the lower-layer FIFO queue that is large enough to allow amortizing the lower-layer
header overhead across multiple transport PDUs. These designs worked header overhead across multiple transport PDUs. These designs worked
well with the Cubic congestion control algorithm, yet the newer well with the cubic congestion control algorithm, yet the newer
generation of CCAs can operate on much smaller queues. To fully generation of algorithms can operate on much smaller queues. To
support latencies less than 1 ms, the home router needs to work fully support latencies less than 1 ms, the home router needs to work
efficiently on sequential transmissions of just a few segments vs. efficiently on sequential transmissions of just a few segments vs.
being optimized for large packet bursts. being optimized for large packet bursts.
Another design trait common in home routers is the use of packet Another design trait common in home routers is the use of packet
aggregation to further amortize the overhead added by the lower-layer aggregation to further amortize the overhead added by the lower-layer
headers. Specifically, multiple IP datagrams are combined into a headers. Specifically, multiple IP datagrams are combined into a
single, large transfer frame. However, this aggregation can add up single, large transfer frame. However, this aggregation can add up
to 10 ms to the packet sojourn delay. to 10 ms to the packet sojourn delay.
Following the famous "you can't improve what you don't measure" Following the famous "you can't improve what you don't measure"
adage, it is important to expose these aggregation delays in a way adage, it is important to expose these aggregation delays in a way
that would allow identifying the source of the bottlenecks and making that would allow identifying the source of the bottlenecks and making
hardware more suitable for the next generation of transport hardware more suitable for the next generation of transport
protocols. protocols.
4.3.6. Cross-Layer Key Points 4.3.6. Cross-Layer Key Points
* Significant differences exist in the characteristics of metrics to * Significant differences exist in the characteristics of metrics to
measured and required optimizations needed in wireless vs. wired be measured and the required optimizations needed in wireless vs.
networks. wired networks.
* Identification of an issue's root cause is hampered by the * Identification of an issue's root cause is hampered by the
challenges in measuring multi-segment network paths. challenges in measuring multi-segment network paths.
* No single component of a network connection has all the data * No single component of a network connection has all the data
required to measure the effects of the complete network required to measure the effects of the complete network
performance on the quality of the end-user experience. performance on the quality of the end-user experience.
* Actionable results require both proper collection and * Actionable results require both proper collection and
interpretation. interpretation.
skipping to change at line 1146 skipping to change at line 1146
types of applications, such as video streaming, file sharing, multi- types of applications, such as video streaming, file sharing, multi-
user gaming, and real-time voice communications. It may be that user gaming, and real-time voice communications. It may be that
asking users for what trade-offs they are willing to accept would be asking users for what trade-offs they are willing to accept would be
a helpful approach: would they rather have a network with low latency a helpful approach: would they rather have a network with low latency
or a network with higher bandwidth? Gamers may make different or a network with higher bandwidth? Gamers may make different
decisions than home office users or content producers, for example. decisions than home office users or content producers, for example.
Furthermore, how can users make these trade-offs in a fair manner Furthermore, how can users make these trade-offs in a fair manner
that does not impact other users? There is a tension between that does not impact other users? There is a tension between
solutions in this space vs. the cost associated with solving these solutions in this space vs. the cost associated with solving these
solutions, as well as which customers are willing to front these problems, as well as which customers are willing to front these
improvement costs. improvement costs.
Challenges in providing higher-priority traffic to users centers Challenges in providing higher-priority traffic to users centers
around the ability for networks to be willing to listen to client around the ability for networks to be willing to listen to client
requests for higher incentives, even though commercial interests may requests for higher incentives, even though commercial interests may
not flow to them without a cost incentive. Shared mediums in general not flow to them without a cost incentive. Shared mediums in general
are subject to oversubscribing, such that the number of users a are subject to oversubscribing, such that the number of users a
network can support is either accurate on an underutilized network or network can support is either accurate on an underutilized network or
may assume an average bandwidth or other usage metric that fails to may assume an average bandwidth or other usage metric that fails to
be accurate during utilization spikes. Individual metrics are also be accurate during utilization spikes. Individual metrics are also
 End of changes. 5 change blocks. 
8 lines changed or deleted 8 lines changed or added

This html diff was produced by rfcdiff 1.48.