As per Relevance of the word standards, we have this rfc below:
Network Working Group G.
Request for Comments: 2679 S.
Category: Standards Track M.
Advanced Network &
September 1999
A One-way Delay Metric for
1. Status of this
This document specifies an Internet standards track protocol for
Internet community, and requests discussion and suggestions
improvements. Please refer to the current edition of the "
Official Protocol Standards" (STD 1) for the standardization
and status of this protocol. Distribution of this memo is unlimited
Copyright
Copyright (C) The Internet Society (1999). All Rights Reserved
2.
This memo defines a metric for one-way delay of packets
Internet paths. It builds on notions introduced and discussed in
IPPM Framework document, RFC 2330 [1]; the reader is assumed to
familiar with that document
This memo is intended to be parallel in structure to a
document for Packet Loss ("A One-way Packet Loss Metric for IPPM")
[2].
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in
document are to be interpreted as described in RFC 2119 [6].
Although RFC 2119 was written with protocols in mind, the key
are used in this document for similar reasons. They are used
ensure the results of measurements from two different
are comparable, and to note instances when an implementation
perturb the network
The structure of the memo is as follows
+ A 'singleton' analytic metric, called Type-P-One-way-Delay,
be introduced to measure a single observation of one-way delay
Almes, et al. Standards Track [Page 1]
RFC 2679 A One-way Delay Metric for IPPM September 1999
+ Using this singleton metric, a 'sample', called Type-P-One-way
Delay-Poisson-Stream, will be introduced to measure a sequence
singleton delays measured at times taken from a Poisson process
+ Using this sample, several 'statistics' of the sample will
defined and discussed
This progression from singleton to sample to statistics, with
separation among them, is important
Whenever a technical term from the IPPM Framework document is
used in this memo, it will be tagged with a trailing asterisk.
example, "term*" indicates that "term" is defined in the Framework
2.1. Motivation
One-way delay of a Type-P* packet from a source host* to
destination host is useful for several reasons
+ Some applications do not perform well (or at all) if end-to-
delay between hosts is large relative to some threshold value
+ Erratic variation in delay makes it difficult (or impossible)
support many real-time applications
+ The larger the value of delay, the more difficult it is
transport-layer protocols to sustain high bandwidths
+ The minimum value of this metric provides an indication of
delay due only to propagation and transmission delay
+ The minimum value of this metric provides an indication of
delay that will likely be experienced when the path* traversed
lightly loaded
+ Values of this metric above the minimum provide an indication
the congestion present in the path
The measurement of one-way delay instead of round-trip delay
motivated by the following factors
+ In today's Internet, the path from a source to a destination
be different than the path from the destination back to the
("asymmetric paths"), such that different sequences of routers
used for the forward and reverse paths. Therefore round-
measurements actually measure the performance of two
paths together. Measuring each path independently highlights
performance difference between the two paths which may
Almes, et al. Standards Track [Page 2]
RFC 2679 A One-way Delay Metric for IPPM September 1999
different Internet service providers, and even radically
types of networks (for example, research versus
networks, or ATM versus packet-over-SONET).
+ Even when the two paths are symmetric, they may have
different performance characteristics due to asymmetric queueing
+ Performance of an application may depend mostly on the
in one direction. For example, a file transfer using TCP
depend more on the performance in the direction that data flows
rather than the direction in which acknowledgements travel
+ In quality-of-service (QoS) enabled networks, provisioning in
direction may be radically different than provisioning in
reverse direction, and thus the QoS guarantees differ.
the paths independently allows the verification of
guarantees
It is outside the scope of this document to say precisely how
metrics would be applied to specific problems
2.2. General Issues Regarding
{Comment: the terminology below differs from that defined by ITU-
documents (e.g., G.810, "Definitions and terminology
synchronization networks" and I.356, "B-ISDN ATM layer cell
performance"), but is consistent with the IPPM Framework document
In general, these differences derive from the different backgrounds
the ITU-T documents historically have a telephony origin, while
authors of this document (and the Framework) have a computer
background. Although the terms defined below have no
equivalent in the ITU-T definitions, after our definitions we
provide a rough mapping. However, note one potential confusion:
definition of "clock" is the computer operating systems
denoting a time-of-day clock, while the ITU-T definition of
denotes a frequency reference.}
Whenever a time (i.e., a moment in history) is mentioned here, it
understood to be measured in seconds (and fractions) relative to UTC
As described more fully in the Framework document, there are
distinct, but related notions of clock uncertainty
Almes, et al. Standards Track [Page 3]
RFC 2679 A One-way Delay Metric for IPPM September 1999
synchronization
measures the extent to which two clocks agree on what time
is. For example, the clock on one host might be 5.4 msec
of the clock on a second host. {Comment: A rough ITU-
equivalent is "time error".}
accuracy
measures the extent to which a given clock agrees with UTC
For example, the clock on a host might be 27.1 msec behind UTC
{Comment: A rough ITU-T equivalent is "time error from UTC".}
resolution
measures the precision of a given clock. For example,
clock on an old Unix host might tick only once every 10 msec
and thus have a resolution of only 10 msec. {Comment: A
rough ITU-T equivalent is "sampling period".}
skew
measures the change of accuracy, or of synchronization,
time. For example, the clock on a given host might gain 1.3
msec per hour and thus be 27.1 msec behind UTC at one time
only 25.8 msec an hour later. In this case, we say that
clock of the given host has a skew of 1.3 msec per
relative to UTC, which threatens accuracy. We might also
of the skew of one clock relative to another clock,
threatens synchronization. {Comment: A rough ITU-T
is "time drift".}
3. A Singleton Definition for One-way
3.1. Metric Name
Type-P-One-way-
3.2. Metric Parameters
+ Src, the IP address of a
+ Dst, the IP address of a
+ T, a
Almes, et al. Standards Track [Page 4]
RFC 2679 A One-way Delay Metric for IPPM September 1999
3.3. Metric Units
The value of a Type-P-One-way-Delay is either a real number, or
undefined (informally, infinite) number of seconds
3.4. Definition
For a real number dT, >>the *Type-P-One-way-Delay* from Src to Dst
T is dT<< means that Src sent the first bit of a Type-P packet to
at wire-time* T and that Dst received the last bit of that packet
wire-time T+dT
>>The *Type-P-One-way-Delay* from Src to Dst at T is
(informally, infinite)<< means that Src sent the first bit of
Type-P packet to Dst at wire-time T and that Dst did not receive
packet
Suggestions for what to report along with metric values appear
Section 3.8 after a discussion of the metric, methodologies
measuring the metric, and error analysis
3.5. Discussion
Type-P-One-way-Delay is a relatively simple analytic metric, and
that we believe will afford effective methods of measurement
The following issues are likely to come up in practice
+ Real delay values will be positive. Therefore, it does not
sense to report a negative value as a real delay. However,
individual zero or negative delay value might be useful as part
a stream when trying to discover a distribution of a stream
delay values
+ Since delay values will often be as low as the 100 usec to 10
range, it will be important for Src and Dst to synchronize
closely. GPS systems afford one way to achieve synchronization
within several 10s of usec. Ordinary application of NTP may
synchronization to within several msec, but this depends on
stability and symmetry of delay properties among those NTP
used, and this delay is what we are trying to measure.
combination of some GPS-based NTP servers and a
designed and deployed set of other NTP servers should yield
results, but this is yet to be tested
+ A given methodology will have to include a way to
whether a delay value is infinite or whether it is merely
large (and the packet is yet to arrive at Dst). As noted
Almes, et al. Standards Track [Page 5]
RFC 2679 A One-way Delay Metric for IPPM September 1999
Mahdavi and Paxson [4], simple upper bounds (such as the 255
seconds theoretical upper bound on the lifetimes of IP
[5]) could be used, but good engineering, including
understanding of packet lifetimes, will be needed in practice
{Comment: Note that, for many applications of these metrics,
harm in treating a large delay as infinite might be zero or
small. A TCP data packet, for example, that arrives only
several multiples of the RTT may as well have been lost.}
+ If the packet is duplicated along the path (or paths) so
multiple non-corrupt copies arrive at the destination, then
packet is counted as received, and the first copy to
determines the packet's one-way delay
+ If the packet is fragmented and if, for whatever reason
reassembly does not occur, then the packet will be deemed lost
3.6. Methodologies
As with other Type-P-* metrics, the detailed methodology will
on the Type-P (e.g., protocol number, UDP/TCP port number, size
precedence).
Generally, for a given Type-P, the methodology would proceed
follows
+ Arrange that Src and Dst are synchronized; that is, that they
clocks that are very closely synchronized with each other and
fairly close to the actual time
+ At the Src host, select Src and Dst IP addresses, and form a
packet of Type-P with these addresses. Any 'padding' portion
the packet needed only to make the test packet a given size
be filled with randomized bits to avoid a situation in which
measured delay is lower than it would otherwise be due
compression techniques along the path
+ At the Dst host, arrange to receive the packet
+ At the Src host, place a timestamp in the prepared Type-P packet
and send it towards Dst
+ If the packet arrives within a reasonable period of time, take
timestamp as soon as possible upon the receipt of the packet.
subtracting the two timestamps, an estimate of one-way delay
be computed. Error analysis of a given implementation of
method must take into account the closeness of
between Src and Dst. If the delay between Src's timestamp and
Almes, et al. Standards Track [Page 6]
RFC 2679 A One-way Delay Metric for IPPM September 1999
actual sending of the packet is known, then the estimate could
adjusted by subtracting this amount; uncertainty in this
must be taken into account in error analysis. Similarly, if
delay between the actual receipt of the packet and Dst's
is known, then the estimate could be adjusted by subtracting
amount; uncertainty in this value must be taken into account
error analysis. See the next section, "Errors and Uncertainties",
for a more detailed discussion
+ If the packet fails to arrive within a reasonable period of time
the one-way delay is taken to be undefined (informally, infinite).
Note that the threshold of 'reasonable' is a parameter of
methodology
Issues such as the packet format, the means by which Dst knows
to expect the test packet, and the means by which Src and Dst
synchronized are outside the scope of this document. {Comment:
plan to document elsewhere our own work in describing such
detailed implementation techniques and we encourage others to
well.}
3.7. Errors and Uncertainties
The description of any specific measurement method should include
accounting and analysis of various sources of error or uncertainty
The Framework document provides general guidance on this point,
we note here the following specifics related to delay metrics
+ Errors or uncertainties due to uncertainties in the clocks of
Src and Dst hosts
+ Errors or uncertainties due to the difference between 'wire time
and 'host time'.
In addition, the loss threshold may affect the results. Each
these are discussed in more detail below, along with a
("Calibration") on accounting for these errors and uncertainties
3.7.1. Errors or uncertainties related to
The uncertainty in a measurement of one-way delay is related,
part, to uncertainties in the clocks of the Src and Dst hosts.
the following, we refer to the clock used to measure when the
was sent from Src as the source clock, we refer to the clock used
measure when the packet was received by Dst as the destination clock
we refer to the observed time when the packet was sent by the
clock as Tsource, and the observed time when the packet was
by the destination clock as Tdest. Alluding to the notions
Almes, et al. Standards Track [Page 7]
RFC 2679 A One-way Delay Metric for IPPM September 1999
synchronization, accuracy, resolution, and skew mentioned in
Introduction, we note the following
+ Any error in the synchronization between the source clock and
destination clock will contribute to error in the
measurement. We say that the source clock and the
clock have a synchronization error of Tsynch if the source
is Tsynch ahead of the destination clock. Thus, if we know
value of Tsynch exactly, we could correct for
synchronization by adding Tsynch to the uncorrected value
Tdest-Tsource
+ The accuracy of a clock is important only in identifying the
at which a given delay was measured. Accuracy, per se, has
importance to the accuracy of the measurement of delay.
computing delays, we are interested only in the
between clock values, not the values themselves
+ The resolution of a clock adds to uncertainty about any
measured with it. Thus, if the source clock has a resolution
10 msec, then this adds 10 msec of uncertainty to any time
measured with it. We will denote the resolution of the
clock and the destination clock as Rsource and Rdest
respectively
+ The skew of a clock is not so much an additional issue as it is
realization of the fact that Tsynch is itself a function of time
Thus, if we attempt to measure or to bound Tsynch, this needs
be done periodically. Over some periods of time, this
can be approximated as a linear function plus some higher
terms; in these cases, one option is to use knowledge of
linear component to correct the clock. Using this correction,
residual Tsynch is made smaller, but remains a source
uncertainty that must be accounted for. We use the
Esynch(t) to denote an upper bound on the uncertainty
synchronization. Thus, |Tsynch(t)| <= Esynch(t).
Taking these items together, we note that naive computation Tdest
Tsource will be off by Tsynch(t) +/- (Rsource + Rdest). Using
notion of Esynch(t), we note that these clock-related
introduce a total uncertainty of Esynch(t)+ Rsource + Rdest.
estimate of total clock-related uncertainty should be included in
error/uncertainty analysis of any measurement implementation
Almes, et al. Standards Track [Page 8]
RFC 2679 A One-way Delay Metric for IPPM September 1999
3.7.2. Errors or uncertainties related to Wire-time vs Host-
As we have defined one-way delay, we would like to measure the
between when the test packet leaves the network interface of Src
when it (completely) arrives at the network interface of Dst, and
refer to these as "wire times." If the timings are
performed by software on Src and Dst, however, then this software
only directly measure the time between when Src grabs a
just prior to sending the test packet and when Dst grabs a
just after having received the test packet, and we refer to these
points as "host times".
To the extent that the difference between wire time and host time
accurately known, this knowledge can be used to correct for host
measurements and the corrected value more accurately estimates
desired (wire time) metric
To the extent, however, that the difference between wire time
host time is uncertain, this uncertainty must be accounted for in
analysis of a given measurement method. We denote by Hsource
upper bound on the uncertainty in the difference between wire
and host time on the Src host, and similarly define Hdest for the
host. We then note that these problems introduce a total
of Hsource+Hdest. This estimate of total wire-vs-host
should be included in the error/uncertainty analysis of
measurement implementation
3.7.3.
Generally, the measured values can be decomposed as follows
measured value = true value + systematic error + random
If the systematic error (the constant bias in measured values) can
determined, it can be compensated for in the reported results
reported value = measured value - systematic
reported value = true value + random
The goal of calibration is to determine the systematic and
error generated by the instruments themselves in as much detail
possible. At a minimum, a bound ("e") should be found such that
reported value is in the range (true value - e) to (true value + e
at least 95 percent of the time. We call "e" the calibration
for the measurements. It represents the degree to which the
Almes, et al. Standards Track [Page 9]
RFC 2679 A One-way Delay Metric for IPPM September 1999
produced by the measurement instrument are repeatable; that is,
closely an actual delay of 30 ms is reported as 30 ms. {Comment: 95
percent was chosen because (1) some confidence level is desirable
be able to remove outliers, which will be found in measuring
physical property; (2) a particular confidence level should
specified so that the results of independent implementations can
compared; and (3) even with a prototype user-level implementation
95% was loose enough to exclude outliers.}
From the discussion in the previous two sections, the error
measurements could be bounded by determining all the
uncertainties, and adding them together to
Esynch(t) + Rsource + Rdest + Hsource + Hdest
However, reasonable bounds on both the clock-related
captured by the first three terms and the host-related
captured by the last two terms should be possible by careful
techniques and calibrating the instruments using a known, isolated
network in a lab
For example, the clock-related uncertainties are greatly
through the use of a GPS time source. The sum of Esynch(t) +
+ Rdest is small, and is also bounded for the duration of
measurement because of the global time source
The host-related uncertainties, Hsource + Hdest, could be bounded
connecting two instruments back-to-back with a high-speed serial
or isolated LAN segment. In this case, repeated measurements
measuring the same one-way delay
If the test packets are small, such a network connection has
minimal delay that may be approximated by zero. The measured
therefore contains only systematic and random error in
instrumentation. The "average value" of repeated measurements is
systematic error, and the variation is the random error
One way to compute the systematic error, and the random error to
95% confidence is to repeat the experiment many times - at
hundreds of tests. The systematic error would then be the median
The random error could then be found by removing the systematic
from the measured values. The 95% confidence interval would be
range from the 2.5th percentile to the 97.5th percentile of
deviations from the true value. The calibration error "e" could
be taken to be the largest absolute value of these two numbers,
the clock-related uncertainty. {Comment: as described, this bound
relatively loose since the uncertainties are added, and the
value of the largest deviation is used. As long as the
Almes, et al. Standards Track [Page 10]
RFC 2679 A One-way Delay Metric for IPPM September 1999
value is not a significant fraction of the measured values, it is
reasonable bound. If the resulting value is a significant
of the measured values, then more exact methods will be needed
compute the calibration error.}
Note that random error is a function of measurement load.
example, if many paths will be measured by one instrument, this
increase interrupts, process scheduling, and disk I/O (for example
recording the measurements), all of which may increase the
error in measured singletons. Therefore, in addition to minimal
measurements to find the systematic error, calibration
should be performed with the same measurement load that
instruments will see in the field
We wish to reiterate that this statistical treatment refers to
calibration of the instrument; it is used to "calibrate the
stick" and say how well the meter stick reflects reality
In addition to calibrating the instruments for finite one-way delay
two checks should be made to ensure that packets reported as
were really lost. First, the threshold for loss should be verified
In particular, ensure the "reasonable" threshold is reasonable:
it is very unlikely a packet will arrive after the threshold value
and therefore the number of packets lost over an interval is
sensitive to the error bound on measurements. Second, consider
possibility that a packet arrives at the network interface, but
lost due to congestion on that interface or to other
exhaustion (e.g. buffers) in the instrument
3.8. Reporting the metric
The calibration and context in which the metric is measured MUST
carefully considered, and SHOULD always be reported along with
results. We now present four items to consider: the Type-P of
packets, the threshold of infinite delay (if any), error calibration
and the path traversed by the test packets. This list is
exhaustive; any additional information that could be useful
interpreting applications of the metrics should also be reported
3.8.1. Type-
As noted in the Framework document [1], the value of the metric
depend on the type of IP packets used to make the measurement,
"type-P". The value of Type-P-One-way-Delay could change if
protocol (UDP or TCP), port number, size, or arrangement for
treatment (e.g., IP precedence or RSVP) changes. The exact Type-
used to make the measurements MUST be accurately reported
Almes, et al. Standards Track [Page 11]
RFC 2679 A One-way Delay Metric for IPPM September 1999
3.8.2. Loss
In addition, the threshold (or methodology to distinguish) between
large finite delay and loss MUST be reported
3.8.3. Calibration
+ If the systematic error can be determined, it SHOULD be
from the measured values
+ You SHOULD also report the calibration error, e, such that
true value is the reported value plus or minus e, with 95%
confidence (see the last section.)
+ If possible, the conditions under which a test packet with
delay is reported as lost due to resource exhaustion on
measurement instrument SHOULD be reported
3.8.4.
Finally, the path traversed by the packet SHOULD be reported,
possible. In general it is impractical to know the precise path
given packet takes through the network. The precise path may
known for certain Type-P on short or stable paths. If Type-
includes the record route (or loose-source route) option in the
header, and the path is short enough, and all routers* on the
support record (or loose-source) route, then the path will
precisely recorded. This is impractical because the route must
short enough, many routers do not support (or are not configured for
record route, and use of this feature would often artificially
the performance observed by removing the packet from common-
processing. However, partial information is still valuable context
For example, if a host can choose between two links* (and hence
separate routes from Src to Dst), then the initial link used
valuable context. {Comment: For example, with Merit's NetNow setup
a Src on one NAP can reach a Dst on another NAP by either of
different backbone networks.}
4. A Definition for Samples of One-way
Given the singleton metric Type-P-One-way-Delay, we now define
particular sample of such singletons. The idea of the sample is
select a particular binding of the parameters Src, Dst, and Type-P
then define a sample of values of parameter T. The means
defining the values of T is to select a beginning time T0, a
time Tf, and an average rate lambda, then define a pseudo-
Almes, et al. Standards Track [Page 12]
RFC 2679 A One-way Delay Metric for IPPM September 1999
Poisson process of rate lambda, whose values fall between T0 and Tf
The time interval between successive values of T will then
1/lambda
{Comment: Note that Poisson sampling is only one way of defining
sample. Poisson has the advantage of limiting bias, but
methods of sampling might be appropriate for different situations
We encourage others who find such appropriate cases to use
general framework and submit their sampling method
standardization.}
4.1. Metric Name
Type-P-One-way-Delay-Poisson-
4.2. Metric Parameters
+ Src, the IP address of a
+ Dst, the IP address of a
+ T0, a
+ Tf, a
+ lambda, a rate in reciprocal
4.3. Metric Units
A sequence of pairs; the elements of each pair are
+ T, a time,
+ dT, either a real number or an undefined number of seconds
The values of T in the sequence are monotonic increasing. Note
T would be a valid parameter to Type-P-One-way-Delay, and that
would be a valid value of Type-P-One-way-Delay
4.4. Definition
Given T0, Tf, and lambda, we compute a pseudo-random Poisson
beginning at or before T0, with average arrival rate lambda,
ending at or after Tf. Those time values greater than or equal to T
and less than or equal to Tf are then selected. At each of the
in this process, we obtain the value of Type-P-One-way-Delay at
time. The value of the sample is the sequence made up of
resulting pairs. If there are no such pairs,
Almes, et al. Standards Track [Page 13]
RFC 2679 A One-way Delay Metric for IPPM September 1999
sequence is of length zero and the sample is said to be empty
4.5. Discussion
The reader should be familiar with the in-depth discussion of
sampling in the Framework document [1], which includes methods
compute and verify the pseudo-random Poisson process
We specifically do not constrain the value of lambda, except to
the extremes. If the rate is too large, then the measurement
will perturb the network, and itself cause congestion. If the
is too small, then you might not capture interesting
behavior. {Comment: We expect to document our experiences with,
suggestions for, lambda elsewhere, culminating in a "best
practices" document.}
Since a pseudo-random number sequence is employed, the sequence
times, and hence the value of the sample, is not fully specified
Pseudo-random number generators of good quality will be needed
achieve the desired qualities
The sample is defined in terms of a Poisson process both to avoid
effects of self-synchronization and also capture a sample that
statistically as unbiased as possible. {Comment: there is,
course, no claim that real Internet traffic arrives according to
Poisson arrival process.} The Poisson process is used to
the delay measurements. The test packets will generally not
at Dst according to a Poisson distribution, since they are
by the network
All the singleton Type-P-One-way-Delay metrics in the sequence
have the same values of Src, Dst, and Type-P
Note also that, given one sample that runs from T0 to Tf, and
new time values T0' and Tf' such that T0 <= T0' <= Tf' <= Tf,
subsequence of the given sample whose time values fall between T0'
and Tf' are also a valid Type-P-One-way-Delay-Poisson-Stream sample
4.6. Methodologies
The methodologies follow directly from
+ the selection of specific times, using the specified
arrival process,
+ the methodologies discussion already given for the
Type-P-One-way-Delay metric
Almes, et al. Standards Track [Page 14]
RFC 2679 A One-way Delay Metric for IPPM September 1999
Care must, of course, be given to correctly handle out-of-
arrival of test packets; it is possible that the Src could send
test packet at TS[i], then send a second one (later) at TS[i+1],
while the Dst could receive the second test packet at TR[i+1],
then receive the first one (later) at TR[i].
4.7. Errors and Uncertainties
In addition to sources of errors and uncertainties associated
methods employed to measure the singleton values that make up
sample, care must be given to analyze the accuracy of the
process with respect to the wire-times of the sending of the
packets. Problems with this process could be caused by
things, including problems with the pseudo-random number
used to generate the Poisson arrival process, or with jitter in
value of Hsource (mentioned above as uncertainty in the
delay metric). The Framework document shows how to use
Anderson-Darling test to verify the accuracy of a Poisson
over small time frames. {Comment: The goal is to ensure that
packets are sent "close enough" to a Poisson schedule, and
periodic behavior.}
4.8. Reporting the metric
You MUST report the calibration and context for the
singletons along with the stream. (See "Reporting the metric"
Type-P-One-way-Delay.)
5. Some Statistics Definitions for One-way
Given the sample metric Type-P-One-way-Delay-Poisson-Stream, we
offer several statistics of that sample. These statistics
offered mostly to be illustrative of what could be done
5.1. Type-P-One-way-Delay-
Given a Type-P-One-way-Delay-Poisson-Stream and a percent X
0% and 100%, the Xth percentile of all the dT values in the Stream
In computing this percentile, undefined values are treated
infinitely large. Note that this means that the percentile
thus be undefined (informally, infinite). In addition, the Type-P
One-way-Delay-Percentile is undefined if the sample is empty
Almes, et al. Standards Track [Page 15]
RFC 2679 A One-way Delay Metric for IPPM September 1999
Example: suppose we take a sample and the results are
Stream1 = <
>
Then the 50th percentile would be 110 msec, since 90 msec and 100
msec are smaller and 110 msec and 'undefined' are larger
Note that if the possibility that a packet with finite delay
reported as lost is significant, then a high percentile (90th
95th) might be reported as infinite instead of finite
5.2. Type-P-One-way-Delay-
Given a Type-P-One-way-Delay-Poisson-Stream, the median of all the
values in the Stream. In computing the median, undefined values
treated as infinitely large. As with Type-P-One-way-Delay
Percentile, Type-P-One-way-Delay-Median is undefined if the sample
empty
As noted in the Framework document, the median differs from the 50
percentile only when the sample contains an even number of values,
which case the mean of the two central values is used
Example: suppose we take a sample and the results are
Stream2 = <
>
Then the median would be 105 msec, the mean of 100 msec and 110 msec
the two central values
5.3. Type-P-One-way-Delay-
Given a Type-P-One-way-Delay-Poisson-Stream, the minimum of all
dT values in the Stream. In computing this, undefined values
treated as infinitely large. Note that this means that the
could thus be undefined (informally, infinite) if all the dT
are undefined. In addition, the Type-P-One-way-Delay-Minimum
Almes, et al. Standards Track [Page 16]
RFC 2679 A One-way Delay Metric for IPPM September 1999
undefined if the sample is empty
In the above example, the minimum would be 90 msec
5.4. Type-P-One-way-Delay-Inverse-
Given a Type-P-One-way-Delay-Poisson-Stream and a time
threshold, the fraction of all the dT values in the Stream less
or equal to the threshold. The result could be as low as 0% (if
the dT values exceed threshold) or as high as 100%. Type-P-One-way
Delay-Inverse-Percentile is undefined if the sample is empty
In the above example, the Inverse-Percentile of 103 msec would
50%.
6. Security
Conducting Internet measurements raises both security and
concerns. This memo does not specify an implementation of
metrics, so it does not directly affect the security of the
nor of applications which run on the Internet. However
implementations of these metrics must be mindful of security
privacy concerns
There are two types of security concerns: potential harm caused
the measurements, and potential harm to the measurements.
measurements could cause harm because they are active, and
packets into the network. The measurement parameters MUST
carefully selected so that the measurements inject trivial amounts
additional traffic into the networks they measure. If they
"too much" traffic, they can skew the results of the measurement,
in extreme cases cause congestion and denial of service
The measurements themselves could be harmed by routers
measurement traffic a different priority than "normal" traffic, or
an attacker injecting artificial measurement traffic. If routers
recognize measurement traffic and treat it separately,
measurements will not reflect actual user traffic. If an
injects artificial traffic that is accepted as legitimate, the
rate will be artificially lowered. Therefore, the
methodologies SHOULD include appropriate techniques to reduce
probability measurement traffic can be distinguished from "normal
traffic. Authentication techniques, such as digital signatures,
be used where appropriate to guard against injected traffic attacks
The privacy concerns of network measurement are limited by the
measurements described in this memo. Unlike passive measurements
there can be no release of existing user data
Almes, et al. Standards Track [Page 17]
RFC 2679 A One-way Delay Metric for IPPM September 1999
7.
Special thanks are due to Vern Paxson of Lawrence Berkeley Labs
his helpful comments on issues of clock uncertainty and statistics
Thanks also to Garry Couch, Will Leland, Andy Scherrer, Sean Shapira
and Roland Wittig for several useful suggestions
8.
[1] Paxson, V., Almes, G., Mahdavi, J. and M. Mathis, "Framework
IP Performance Metrics", RFC 2330, May 1998.
[2] Almes, G., Kalidindi, S. and M. Zekauskas, "A One-way
Loss Metric for IPPM", RFC 2680, September 1999.
[3] Mills, D., "Network Time Protocol (v3)", RFC 1305, April 1992.
[4] Mahdavi J. and V. Paxson, "IPPM Metrics for
Connectivity", RFC 2678, September 1999.
[5] Postel, J., "Internet Protocol", STD 5, RFC 791, September 1981.
[6] Bradner, S., "Key words for use in RFCs to Indicate
Levels", BCP 14, RFC 2119, March 1997.
[7] Bradner, S., "The Internet Standards Process -- Revision 3",
9, RFC 2026, October 1996.
Almes, et al. Standards Track [Page 18]
RFC 2679 A One-way Delay Metric for IPPM September 1999
9. Authors'
Guy
Advanced Network & Services, Inc
200 Business Park
Armonk, NY 10504
Phone: +1 914 765 1120
EMail: almes@advanced.
Sunil
Advanced Network & Services, Inc
200 Business Park
Armonk, NY 10504
Phone: +1 914 765 1128
EMail: kalidindi@advanced.
Matthew J.
Advanced Network & Services, Inc
200 Business Park
Armonk, NY 10504
Phone: +1 914 765 1112
EMail: matt@advanced.
Almes, et al. Standards Track [Page 19]
RFC 2679 A One-way Delay Metric for IPPM September 1999
10. Full Copyright
Copyright (C) The Internet Society (1999). All Rights Reserved
This document and translations of it may be copied and furnished
others, and derivative works that comment on or otherwise explain
or assist in its implementation may be prepared, copied,
and distributed, in whole or in part, without restriction of
kind, provided that the above copyright notice and this paragraph
included on all such copies and derivative works. However,
document itself may not be modified in any way, such as by
the copyright notice or references to the Internet Society or
Internet organizations, except as needed for the purpose
developing Internet standards in which case the procedures
copyrights defined in the Internet Standards process must
followed, or as required to translate it into languages other
English
The limited permissions granted above are perpetual and will not
revoked by the Internet Society or its successors or assigns
This document and the information contained herein is provided on
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED,
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE
Funding for the RFC Editor function is currently provided by
Internet Society
Almes, et al. Standards Track [Page 20]
if you see any problems within the linking, don't worry be happy,
this is version 0.1 of the Relevance System and you gotta expect some crappy subroutines sometimes,
just be content we did not write this in Java, which would have made this "bigger and better" HAHAHHA.
RFC documents can be found at I.E.T.F.
Relevance System Copyright © 2002 Spectrum WorldResearch
other technical nosh by ServerMasters Corporation
collaboration of BobX
