rfc8889xml2.original.xml   rfc8889.xml 
<?xml version="1.0" encoding="UTF-8"?> <?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!-- A set of on-line citation libraries are maintained on the xml2rfc web site.
The next line defines an entity named RFC2629, which contains the necessary
XML
for the reference element, and is used much later in the file. This XML co
ntains an
anchor (also RFC2629) which can be used to cross-reference this item in the
text.
You can also use local file names instead of a URI. The environment variab
le
XML_LIBRARY provides a search path of directories to look at to locate a
relative path name for the file. There has to be one entity for each item t
o be
referenced. -->
<!ENTITY RFC2234 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC
.2234.xml">
<!ENTITY RFC2629 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC
.2629.xml">
<!ENTITY RFC4234 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC
.4234.xml">
<!ENTITY RFC5575 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC
.5575.xml">
<!-- There is also a library of current Internet Draft citations. It isn't a go
od idea to
actually use one for the template because it might have disappeared when yo
u come to test
this template. This is the form of the entity definition
&lt;!ENTITY I-D.mrose-writing-rfcs SYSTEM
"http://xml.resource.org/public/rfc/bibxml3/reference.I-D.mrose-writing-rfc
s.xml">
corresponding to a draft filename draft-mrose-writing-rfcs-nn.txt. The cita
tion will be
to the most recent draft in the sequence, and is updated roughly hourly on
the web site.
For working group drafts, the same principle applies: file name starts draf
t-ietf-wgname-..
and entity file is reference.I-D.ietf-wgname-... The corresponding entity
name is
I-D.ietf-wgname-... (I-D.mrose-writing-rfcs for the other example). Of cou
rse this doesn't
change when the draft version changes.
-->
<!-- Fudge for XMLmind which doesn't have this built in -->
<!ENTITY nbsp "&#160;">
]>
<!-- Extra statement used by XSLT processors to control the output style. -->
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<!-- Processing Instructions can be placed here but if you are editing
with XMLmind (and maybe other XML editors) they are better placed
after the rfc element start tag as shown below. -->
<!-- Information about the document. <!DOCTYPE rfc SYSTEM "rfc2629-xhtml.ent">
category values: std, bcp, info, exp, and historic
For Internet-Drafts, specify attribute "ipr".
(ipr values are: full3667, noModification3667, noDerivatives3667),
Also for Internet-Drafts, can specify values for
attributes "docName" and, if relevant, "iprExtract". Note
that the value for iprExtract is the anchor attribute
value of a section (such as a MIB specification) that can be
extracted for separate publication, and is only
useful whenhe value of "ipr" is not "full3667". -->
<!-- TODO: verify which attributes are specified only
by the RFC editor. It appears that attributes
"number", "obsoletes", "updates", and "seriesNo"
are specified by the RFC editor (and not by
the document author). -->
<rfc
category="exp"
ipr="trust200902"
docName="draft-ietf-ippm-multipoint-alt-mark-09" >
<!-- Processing Instructions- PIs (for a complete list and description,
see file http://xml.resource.org/authoring/README.html and below... --
>
<!-- Some of the more generally applicable PIs that most I-Ds might want to
use -->
<!-- Try to enforce the ID-nits conventions and DTD validity -->
<?rfc strict="yes" ?>
<!-- Items used when reviewing the document -->
<?rfc comments="no" ?> <!-- Controls display of <cref> elements -->
<?rfc inline="no" ?> <!-- When no, put comments at end in comments sectio
n,
otherwise, put inline -->
<?rfc editing="no" ?> <!-- When yes, insert editing marks: editing marks c
onsist of a
string such as <29> printed in the blank line a
t the
beginning of each paragraph of text. -->
<!-- Create Table of Contents (ToC) and set some options for it.
Note the ToC may be omitted for very short documents,but idnits insists
on a ToC
if the document has more than 15 pages. -->
<?rfc toc="yes"?>
<?rfc tocompact="yes"?> <!-- If "yes" eliminates blank lines before main sect
ion entries. -->
<?rfc tocdepth="3"?> <!-- Sets the number of levels of sections/subsection
s... in ToC -->
<!-- Choose the options for the references.
Some like symbolic tags in the references (and citations) and others pr
efer
numbers. The RFC Editor always uses symbolic tags.
The tags used are the anchor attributes of the references. -->
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes" ?> <!-- If "yes", causes the references to be sorted in
order of tags.
This doesn't have any effect unless symrefs is
"yes" also. -->
<!-- These two save paper: Just setting compact to "yes" makes savings by no
t starting each
main section on a new page but does not omit the blank lines between li
st items.
If subcompact is also "yes" the blank lines between list items are also
omitted. -->
<?rfc compact="yes" ?>
<?rfc subcompact="no" ?>
<!-- end of list of popular I-D processing instructions -->
<!-- ***** FRONT MATTER ***** --> <rfc xmlns:xi="http://www.w3.org/2001/XInclude" ipr="trust200902"
docName="draft-ietf-ippm-multipoint-alt-mark-09" number="8889"
obsoletes="" updates="" submissionType="IETF" category="exp"
consensus="true" xml:lang="en" tocInclude="true" tocDepth="3"
symRefs="true" sortRefs="true" version="3">
<front> <front>
<!-- The abbreviated title is used in the page header - it is only necessary <title abbrev="Multipoint AM">Multipoint Alternate-Marking Method for
if the Passive and Hybrid Performance Monitoring</title>
full title is longer than 42 characters --> <seriesInfo name="RFC" value="8889"/>
<title abbrev="Multipoint AM">Multipoint Alternate Marking method for passiv <author role="editor" fullname="Giuseppe Fioccola" initials="G." surname=
e and hybrid performance monitoring</title> "Fioccola">
<!-- add 'role="editor"' below for the editors if appropriate -->
<author
role="editor"
fullname="Giuseppe Fioccola"
initials="G."
surname="Fioccola">
<!-- abbrev not needed but can be used for the header
if the full organization name is too long -->
<organization>Huawei Technologies</organization> <organization>Huawei Technologies</organization>
<address> <address>
<postal> <postal>
<street>Riesstrasse, 25</street> <street>Riesstrasse, 25</street>
<city>Munich</city> <city>Munich</city>
<code>80992</code> <code>80992</code>
<country>Germany</country> <country>Germany</country>
</postal> </postal>
<email>giuseppe.fioccola@huawei.com</email> <email>giuseppe.fioccola@huawei.com</email>
<!--
If I had a phone, fax machine, and a URI, I could add the following:
<phone>+1-408-555-1234</phone>
<facsimile>+1-555-911-9111</facsimile>
<uri>http://www.example.com/</uri>
-->
</address> </address>
</author> </author>
<author fullname="Mauro Cociglio" initials="M." surname="Cociglio">
<author
fullname="Mauro Cociglio"
initials="M."
surname="Cociglio">
<!-- abbrev not needed but can be used for the header
if the full organization name is too long -->
<organization>Telecom Italia</organization> <organization>Telecom Italia</organization>
<address> <address>
<postal> <postal>
<street>Via Reiss Romoli, 274</street> <street>Via Reiss Romoli, 274</street>
<city>Torino</city> <city>Torino</city>
<code>10148</code> <code>10148</code>
<country>Italy</country> <country>Italy</country>
</postal> </postal>
<email>mauro.cociglio@telecomitalia.it</email> <email>mauro.cociglio@telecomitalia.it</email>
<!--
If I had a phone, fax machine, and a URI, I could add the following:
<phone>+1-408-555-1234</phone>
<facsimile>+1-555-911-9111</facsimile>
<uri>http://www.example.com/</uri>
-->
</address> </address>
</author> </author>
<author fullname="Amedeo Sapio" initials="A." surname="Sapio">
<author <organization>Intel Corporation</organization>
fullname="Amedeo Sapio" <address>
initials="A." <postal>
surname="Sapio"> <street>4750 Patrick Henry Dr.</street>
<!-- abbrev not needed but can be used for the header <city>Santa Clara</city>
if the full organization name is too long --> <region>CA</region>
<organization>Politecnico di Torino</organization> <code>95054</code>
<address> <country>USA</country>
<postal> </postal>
<street>Corso Duca degli Abruzzi, 24</street> <email>amedeo.sapio@intel.com</email>
<city>Torino</city>
<code>10129</code>
<country>Italy</country>
</postal>
<email>amedeo.sapio@polito.it</email>
<!--
If I had a phone, fax machine, and a URI, I could add the following:
<phone>+1-408-555-1234</phone>
<facsimile>+1-555-911-9111</facsimile>
<uri>http://www.example.com/</uri>
-->
</address> </address>
</author> </author>
<author fullname="Riccardo Sisto" initials="R." surname="Sisto">
<author
fullname="Riccardo Sisto"
initials="R."
surname="Sisto">
<!-- abbrev not needed but can be used for the header
if the full organization name is too long -->
<organization>Politecnico di Torino</organization> <organization>Politecnico di Torino</organization>
<address> <address>
<postal> <postal>
<street>Corso Duca degli Abruzzi, 24</street> <street>Corso Duca degli Abruzzi, 24</street>
<city>Torino</city> <city>Torino</city>
<code>10129</code> <code>10129</code>
<country>Italy</country> <country>Italy</country>
</postal> </postal>
<email>riccardo.sisto@polito.it</email> <email>riccardo.sisto@polito.it</email>
<!--
If I had a phone, fax machine, and a URI, I could add the following:
<phone>+1-408-555-1234</phone>
<facsimile>+1-555-911-9111</facsimile>
<uri>http://www.example.com/</uri>
-->
</address> </address>
</author> </author>
<date year="2020" month="August" />
<date year="2020"/> <!-- month="March" is no longer necessary <area/>
note also, day="30" is optional -->
<!-- WARNING: If the month and year are the current ones, xml2rfc will fill
in the day for
you. If only the year is specified, xml2rfc will fill in the current da
y and month
irrespective of the day. This silliness should be fixed in v1.31. -->
<!-- Meta-data Declarations -->
<!-- Notice the use of &amp; as an escape for & which would otherwise
start an entity declaration, whereas we want a literal &. -->
<area></area>
<!-- WG name at the upperleft corner of the doc,
IETF fine for individual submissions. You can also
omit this element in which case in defaults to "Network Working Group"
-
a hangover from the ancient history of the IETF! -->
<workgroup>IPPM Working Group</workgroup> <workgroup>IPPM Working Group</workgroup>
<!-- The DTD allows multiple area and workgroup elements but only the first
one has any
effect on output. -->
<!-- You can add <keyword/> elements here. They will be incorporated into H
TML output
files in a meta tag but they have no effect on text or nroff output. --
>
<abstract> <abstract>
<t>The Alternate Marking method, as presented in RFC 8321, <t>The Alternate-Marking method, as presented in RFC 8321,
can be applied only to point-to-point flows because it assumes th can only be applied to point-to-point flows, because it assumes t
at all the packets hat all the packets
of the flow measured on one node are measured again by a single s econd node. of the flow measured on one node are measured again by a single s econd node.
This document generalizes and expands this methodology to measure any kind of This document generalizes and expands this methodology to measure any kind of
unicast flows, whose packets can follow several different paths i unicast flow whose packets can follow several different paths in
n the the
network, in wider terms a multipoint-to-multipoint network. For network -- in wider terms, a multipoint-to-multipoint network. F
this reason or this reason,
the technique here described is called Multipoint Alternate Marki the technique here described is called "Multipoint Alternate Mark
ng.</t> ing".</t>
</abstract> </abstract>
</front>
<middle>
<section numbered="true" toc="default">
<name>Introduction</name>
<t>The Alternate-Marking method, as described in <xref target="RFC8321" fo
rmat="default">RFC 8321</xref>, is applicable to a point-to-point path.
The extension proposed in this document applies to the most general case o
f multipoint-to-multipoint path
and enables flexible and adaptive performance measurements in a managed ne
twork.</t>
<t>The Alternate-Marking methodology described in <xref target="RFC8321" f
ormat="default">RFC 8321</xref>
allows the synchronization of the measurements in different points by divi
ding the packet flow
into batches. So it is possible to get coherent counters and show what is
happening in every
marking period for each monitored flow. The monitoring parameters are the
packet counter and timestamps
of a flow for each marking period. Note that additional details about the
applicability of the
Alternate-Marking methodology are described in <xref
target="RFC8321" format="default">RFC 8321</xref> while implementation det
ails can be found in the
paper "AM-PM: Efficient Network Telemetry using Alternate Marking" <xref
target="IEEE-Network-PNPM" format="default"/>.</t>
<t>There are some applications of the Alternate-Marking method where there
are a lot of
monitored flows and nodes. Multipoint Alternate Marking aims to reduce the
se values and
makes the performance monitoring more flexible in case a detailed analysis
is not needed.
For instance, by considering n measurement points and m monitored flows,
the order of magnitude
of the packet counters for each time interval is n*m*2 (1 per color). The
number of
measurement points and monitored flows may vary and depends on the portion
of the network
we are monitoring (core network, metro network, access network) and the gr
anularity
(for each service, each customer). So if both n and m are high values, the
packet counters
increase a lot, and Multipoint Alternate Marking offers a tool to control
these parameters.</t>
<t>The approach presented in this document is applied only to unicast flow
s and not to multicast.
Broadcast, Unknown Unicast, and Multicast (BUM) traffic is not
considered here, because traffic replication
is not covered by the Multipoint Alternate-Marking method. Furthermore,
it can be applicable to anycast flows, and Equal-Cost Multipath (ECMP)
paths can also be easily monitored with this technique.</t>
<t>In short, <xref target="RFC8321" format="default">RFC 8321</xref>
applies to point-to-point unicast flows and BUM traffic, while this
document and its Clustered Alternate-Marking method is valid for
multipoint-to-multipoint
unicast flows, anycast, and ECMP flows.</t>
<t>Therefore,the Alternate-Marking method can be extended to any kind of
multipoint-to-multipoint paths,
and the network-clustering approach presented in this document is the form
alization of how to
implement this property and allow a flexible and optimized performance mea
surement support
for network management in every situation.</t>
<t>Without network clustering, it is possible to apply Alternate Marking o
nly for all
the network or per single flow. Instead, with network clustering, it is po
ssible to use the partition
of the network into clusters at different levels in order to perform the n
eeded degree of detail.
In some circumstances, it is possible to monitor a multipoint network by a
nalyzing the network clustering,
without examining in depth. In case of problems (packet loss is measured o
r the delay is too high),
the filtering criteria could be specified more in order to perform a detai
led analysis by using a
different combination of clusters up to a per-flow measurement as describe
d in <xref target="RFC8321" format="default">RFC 8321</xref>.</t>
<t>This approach fits very well with the Closed-Loop Network and Software-
Defined Network (SDN) paradigm,
where the SDN orchestrator and the SDN controllers are the brains of the n
etwork and can manage flow control
to the switches and routers and, in the same way, can calibrate the perfor
mance measurements
depending on the desired accuracy. An SDN controller application can orche
strate how accurately the network
performance monitoring is set up by applying the Multipoint Alternate Mark
ing as described in this document.</t>
<t>It is important to underline that, as an extension of <xref target="RFC
8321" format="default">RFC 8321</xref>, this is a methodology document,
so the mechanism that can be used to transmit the counters and the timesta
mps is out of scope here, and the implementation
is open. Several options are possible -- e.g., see "Enhanced Alternate Mar
king Method" <xref target="I-D.zhou-ippm-enhanced-alternate-marking" format="def
ault"/>.</t>
</front> <t>
Note that the fragmented packets case can be managed with the
<middle> Alternate-Marking methodology only if fragmentation happens outside
<section title="Introduction"> the portion of the network that is monitored. This is always true for both <xref
<t>The Alternate Marking method, as described in <xref target="RFC8321"> target="RFC8321" format="default">RFC 8321</xref> and Multipoint Alternate Mark
RFC 8321</xref>, is applicable to a point-to-point path. ing, as explained here.
The extension proposed in this document applies to the most gener </t>
al case of multipoint-to-multipoint path
and enables flexible and adaptive performance measurements in a m
anaged network.</t>
<t>The Alternate Marking methodology described in <xref target="R
FC8321">RFC 8321</xref>
allows the synchronization of the measurements in different point
s by dividing the packet flow
into batches. So it is possible to get coherent counters and show
what is happening in every
marking period for each monitored flow. The monitoring parameters
are the packet counter and timestamps
of a flow for each marking period. Note that additional details a
bout the applicability of the
Alternate Marking methodology are described both in <xref target=
"RFC8321">RFC 8321</xref> and
in the paper <xref target="IEEE-Network-PNPM"/>.</t>
<t>There are some applications of the Alternate Marking method wh
ere there are a lot of
monitored flows and nodes. Multipoint Alternate Marking aims to r
educe these values and
makes the performance monitoring more flexible in case a detailed
analysis is not needed.
For instance, by considering n measurement points and m monitored
flows,the order of magnitude
of the packet counters for each time interval is n*m*2 (1 per col
or). The number of
measurement points and monitored flows may vary and depends on th
e portion of the network
we are monitoring (core network, metro network, access network) a
nd on the granularity
(for each service, each customer). So if both n and m are high va
lues the packet counters
increase a lot and Multipoint Alternate Marking offers a tool to
control these parameters.</t>
<t>The approach presented in this document is applied only to uni
cast flows and not to multicast.
Broadcast, Unknown-unicast, and Multicast (BUM) traffic is not co
nsidered here, because traffic replication
is not covered by the Multipoint Alternate Marking method. Furthe
rmore it can be applicable to anycast
flows and Equal-Cost MultiPath (ECMP) paths can also be easily mo
nitored with this technique.</t>
<t>In short, <xref target="RFC8321">RFC 8321</xref> applies to po
int-to-point unicast flows and BUM
traffic while this document and its Clustered Alternate Marking m
ethod is valid for multipoint-to-multipoint
unicast flows, anycast and ECMP flows.</t>
<t>The Alternate Marking method can therefore be extended to any
kind of multipoint to multipoint paths,
and the network clustering approach presented in this document is
the formalization of how to
implement this property and allow a flexible and optimized perfor
mance measurement support
for network management in every situation.</t>
<t>Without network clustering, it is possible to apply Alternate
Marking only for all
the network or per single flow. Instead, with network clustering,
it is possible to use the partition
of the network into clusters at different levels in order to perf
orm the needed degree of detail.
In some circumstances it is possible to monitor a Multipoint Netw
ork by analysing the Network Clustering,
without examining in depth. In case of problems (packet loss is m
easured or the delay is too high)
the filtering criteria could be specified more in order to perfor
m a detailed analysis by using a
different combination of clusters up to a per-flow measurement as
described in <xref target="RFC8321">RFC 8321</xref>.</t>
<t>This approach fits very well with the Closed Loop Network and
Software Defined Network (SDN) paradigm
where the SDN Orchestrator and the SDN Controllers are the brains
of the network and can manage flow control
to the switches and routers and, in the same way, can calibrate t
he performance measurements
depending on the desired accuracy. An SDN Controller Application
can orchestrate how accurate the network
performance monitoring is setup by applying the Multipoint Altern
ate Marking as described in this document.</t>
<t>It is important to underline that, as extension of <xref targe
t="RFC8321">RFC 8321</xref>, this is a methodology draft,
so the mechanism that can be used to transmit the counters and th
e timestamps is out of scope here and the implementation
is open. Several options are possible, e.g. <xref target="I-D.zho
u-ippm-enhanced-alternate-marking"/>.</t>
<t>Note that, as for <xref target="RFC8321">RFC 8321</xref>, the
fragmented packets case can be managed with this methodology
if fragmentation happens outside the portion of the monitored net
work.</t>
</section> </section>
<section numbered="true" toc="default">
<name>Terminology</name>
<t>The definitions of the basic terms are identical to those found in
Alternate Marking <xref target="RFC8321" format="default"/>.
It is to be remembered that <xref target="RFC8321" format="default">RFC 8
321</xref> is valid for point-to-point unicast flows and BUM traffic.</t>
<t>The important new terms that need to be explained are listed below:</t>
<section title="Terminology"> <dl>
<dt>Multipoint Alternate Marking:</dt><dd>Extension to <xref target="RFC8321"
<t>The definitions of the basic terms are identical to those found in Alt format="default">RFC 8321</xref>, valid for multipoint-to-multipoint
ernate Marking (<xref target="RFC8321">RFC 8321</xref>). unicast flows, anycast, and ECMP flows. It can also be referred to as Clustere
It is to be remembered that <xref target="RFC8321">RFC 8321</xref> is val d
id for point-to-point unicast flows and BUM traffic.</t> Alternate Marking.</dd>
<dt>Flow definition:</dt><dd>The concept of flow is generalized in this
<t>The important new terms that need to be explained are listed below:<li document. The identification fields are selected without any constraints
st> and, in general, the flow can be a multipoint-to-multipoint flow, as a
result of aggregate point-to-point flows.</dd>
<t>Multipoint Alternate Marking: Extension to <xref target="RFC8321">RFC <dt>Monitoring network:</dt><dd>Identified with the nodes of the
8321</xref>, valid for multipoint-to-multipoint network that are the measurement points (MPs) and
unicast flows, anycast and ECMP flows. It can also be referred as Cluster the links that are the connections between MPs. The monitoring network graph
ed Alternate Marking;</t> depends on the flow definition, so it can represent a specific flow or the
entire network topology as aggregate of all the flows.</dd>
<t>Flow definition: The concept of flow is generalized in this document. <dt>Cluster:</dt><dd>Smallest identifiable subnetwork of the entire
The identification fields are selected without any constraints monitoring network graph that still satisfies the condition that the number
and, in general, the flow can be a multipoint-to-multipoint flow, as a re of packets that go in is the same as the number that go out.</dd>
sult of aggregate point-to-point flows;</t> <dt>Multipoint metrics:</dt><dd>Packet loss, delay, and delay variation are
extended to the case of multipoint flows.
<t>Monitoring Network: it is identified with the nodes of the network tha It is possible to compute these metrics on the basis of multipoint paths in or
t are the measurement points (MPs) and der
the links that are the connections between MPs. The Monitoring Network gr to associate the measurements to a cluster, a combination of clusters, or
aph depends on the flow definition, so it can the entire monitored network. For delay and delay variation,
represent a specific flow or the the entire network topology as aggregate it is also possible to define the metrics on a single-packet basis, and it
of all the flows;</t> means that the multipoint path is used to easily couple packets between
input and output nodes of a multipoint path.</dd>
<t>Cluster: smallest identifiable subnetwork of the entire Monitoring Net </dl>
work graph that still
satisfies the condition that the number of packets that goes in is the sa
me that goes out;</t>
<t>Multipoint metrics: packet loss, delay and delay variation are extende
d to the case of multipoint flows.
It is possible to compute these metrics on multipoint paths basis in orde
r to associate the measurements
to a cluster, to a combination of clusters or to the entire monitored net
work. For delay and delay variation,
it is also possible to define the metrics on a single packet basis and it
means that the multipoint path is used
to easily couple packets between input and output nodes of a multipoint p
ath.</t>
</list></t>
<t>The next section highlights the correlation with the terms used in <xr
ef target="RFC5644">RFC 5644</xref>.</t>
<section title="Correlation with RFC5644">
<t><xref target="RFC5644">RFC 5644</xref> is limited to active measureme
nts
using a single source packet or stream, and observations of corresponding p
ackets along the
path (spatial), at one or more destinations (one-to-group), or both.</t>
<t>Instead, the scope of this memo is to define multiparty metrics for p <t>The next section highlights the correlation with the terms used in
assive and hybrid <xref target="RFC5644" format="default">RFC 5644</xref>.</t>
<section numbered="true" toc="default">
<name>Correlation with RFC 5644</name>
<t><xref target="RFC5644" format="default">RFC 5644</xref> is limited to
active measurements
using a single source packet or stream. Its scope is also limited to observ
ations of corresponding packets along the
path (spatial metric) and at one or more destinations (one-to-group) along
the path.</t>
<t>Instead, the scope of this memo is to define multiparty metrics for p
assive and hybrid
measurements in a group-to-group topology with multiple sources and dest inations.</t> measurements in a group-to-group topology with multiple sources and dest inations.</t>
<t><xref target="RFC5644" format="default">RFC 5644</xref> introduces
<t><xref target="RFC5644">RFC 5644</xref> introduces metric names that c metric names that can be reused here
an be reused also here but have to be extended and rephrased to be applied to the Alternate-Mar
but have to be extended and rephrased to be applied to the Alternate Mar king schema:</t>
king schema:</t> <ol spacing="normal" type="a">
<t><list style="letters"> <li>the multiparty metrics are not only one-to-group metrics but can b
<t>the multiparty metrics are not only one-to-group metrics but can be also e also group-to-group
group-to-group metrics;</li>
metrics;</t> <li>the spatial metrics, used for measuring the performance of segment
<t>the spatial metrics, used for measuring the performance of segments of a s of a source to destination path,
source to destination path, are applied here to group-to-group segments (called clusters).</li>
are applied here to group-to-group segments (called Clusters).</t> </ol>
</list></t> </section>
</section>
</section> <section numbered="true" toc="default">
</section> <name>Flow Classification</name>
<t>A unicast flow is identified by all the packets having a set of common
<section title="Flow classification"> characteristics.
<t>An unicast flow is identified by all the packets having a set This definition is inspired by <xref target="RFC7011" format="def
of common characteristics. ault">RFC 7011</xref>.</t>
This definition is inspired by <xref target="RFC7011">RFC 7011</x <t>As an example, by considering a flow as all the packets sharing the sam
ref>.</t> e
<t>As an example, by considering a flow as all the packets sharin
g the same
source IP address or the same destination IP address, it is easy to understand source IP address or the same destination IP address, it is easy to understand
that the resulting pattern will not be a point-to-point connectio n, that the resulting pattern will not be a point-to-point connectio n,
but a point-to-multipoint or multipoint-to-point connection.</t> but a point-to-multipoint or multipoint-to-point connection.</t>
<t>In general, a flow can be defined by a set of selection rules used to m
<t>In general a flow can be defined by a set of selection rules u atch
sed to match
a subset of the packets processed by the network device. These ru les a subset of the packets processed by the network device. These ru les
specify a set of layer-3 and layer-4 headers fields (Identificati on Fields) specify a set of Layer 3 and Layer 4 header fields (identificatio n fields)
and the relative values that must be found in matching packets.</ t> and the relative values that must be found in matching packets.</ t>
<t>The choice of the identification fields directly affects the type of
<t>The choice of the identification fields directly affects the t
ype of
paths that the flow would follow in the network. In fact, it is paths that the flow would follow in the network. In fact, it is
possible to relate a set of identification fields with the patter n possible to relate a set of identification fields with the patter n
of the resulting graphs, as listed in Figure 1.</t> of the resulting graphs, as listed in <xref target="Flows" />.</t
>
<t>A TCP 5-tuple usually identifies flows following either a sing <t>A TCP 5-tuple usually identifies flows following either a single path
le path or a point-to-point multipath (in the case of load balancing). On
or a point-to-point multipath (in case of load balancing). On the the contrary,
contrary,
a single source address selects aggregate flows following a point -to-multipoint, a single source address selects aggregate flows following a point -to-multipoint,
while a multipoint-to-point can be the result of a matching on a single while a multipoint-to-point can be the result of a matching on a single
destination address. destination address.
In case a selection rule and its reverse are used for bidirection al measurements, In the case where a selection rule and its reverse are used for b idirectional measurements,
they can correspond to a point-to-multipoint in one direction they can correspond to a point-to-multipoint in one direction
and a multipoint-to-point in the opposite direction.</t> and a multipoint-to-point in the opposite direction.</t>
<t>So the flows to be monitored are selected into the monitoring points
<t>So the flows to be monitored are selected into the monitoring using packet selection rules, which can also change the pattern o
points f the monitored
using packet selection rules, that can also change the pattern of
the monitored
network.</t> network.</t>
<t>Note that, more in general, the flow can be defined at differe <t>Note that, more generally, the flow can be defined at different levels
nt levels based based
on the encapsulation considered and additional conditions that ar on the potential encapsulation, and additional conditions that ar
e not in the packet header e not in the packet header
can also be included as part of matching criteria.</t> can also be included as part of matching criteria.</t>
<t>The Alternate-Marking method is applicable only to a single path (and
<t>The Alternate Marking method is applicable only to a single pa
th (and
partially to a one-to-one multipath), so the extension proposed i n partially to a one-to-one multipath), so the extension proposed i n
this document is suitable also for the most general case of multi point-to-multipoint, this document is suitable also for the most general case of multi point-to-multipoint,
which embraces all the other patterns of Figure 1.</t> which embraces all the other patterns of <xref target="Flows" />.
</t>
<figure anchor="Flows" title="Flow classification"> <figure anchor="Flows">
<artwork><![CDATA[ <name>Flow Classification</name>
<artwork name="" type="" align="left" alt=""><![CDATA[
point-to-point single path point-to-point single path
+------+ +------+ +------+ +------+ +------+ +------+
---<> R1 <>----<> R2 <>----<> R3 <>--- ---<> R1 <>----<> R2 <>----<> R3 <>---
+------+ +------+ +------+ +------+ +------+ +------+
point-to-point multipath point-to-point multipath
+------+ +------+
<> R2 <> <> R2 <>
/ +------+ \ / +------+ \
skipping to change at line 459 skipping to change at line 302
+------+ \ +------+ \
+------+ \ +------+ +------+ \ +------+
---<> R2 <> <> R7 <>--- ---<> R2 <> <> R7 <>---
+------+ \ / +------+ +------+ \ / +------+
\ +------+ / \ +------+ /
<> R5 <> <> R5 <>
/ +------+ \ / +------+ \
+------+ / \ +------+ +------+ / \ +------+
---<> R3 <> <> R8 <>--- ---<> R3 <> <> R8 <>---
+------+ +------+ +------+ +------+
]]></artwork> ]]></artwork>
</figure> </figure>
<t>The case of unicast flow is considered in <xref target="Flows"/>. The a
<t>The case of unicast flow is considered in the previous figure. nycast flow
Anyway the anycast flow is also in scope, because there is no replication and only a single node
is also in scope because there is no replication and only a singl from the anycast group
e node from the anycast group receives the traffic, so it can be viewed as a special case of unicast flo
receives the traffic, so it can be viewed as a special case of un w. Furthermore, an
icast flow. Furthermore, an ECMP flow is in scope by definition, since it is a point-to-multipoint uni
ECMP flow is in scope by definition, since it is a point-to-multi cast flow.</t>
point unicast flow.</t>
</section> </section>
<section numbered="true" toc="default">
<section title="Multipoint Performance Measurement"> <name>Multipoint Performance Measurement</name>
<t>By Using the Alternate Marking method only point-to-point paths can b <t>By using the Alternate-Marking method, only point-to-point paths can be
e monitored. monitored.
To have an IP (TCP/UDP) flow that follows a point-to-point path we have t To have an IP (TCP/UDP) flow that follows a point-to-point path, we have
o define, to define,
with a specific value, 5 identification fields (IP Source, IP Destination , with a specific value, 5 identification fields (IP Source, IP Destination ,
Transport Protocol, Source Port, Destination Port).</t> Transport Protocol, Source Port, Destination Port).</t>
<t>Multipoint Alternate Marking enables the performance measurement for mu
<t>Multipoint Alternate Marking enables the performance measurement for m ltipoint flows
ultipoint flows
selected by identification fields without any constraints (even the entir e network production traffic). selected by identification fields without any constraints (even the entir e network production traffic).
It is also possible to use multiple marking points for the same monitored flow.</t> It is also possible to use multiple marking points for the same monitored flow.</t>
<section numbered="true" toc="default">
<section title="Monitoring Network"> <name>Monitoring Network</name>
<t>The Monitoring Network is deduced from the Production Network, by iden <t>The monitoring network is deduced from the production network by iden
tifying tifying
the nodes of the graph that are the measurement points, and the links tha t are the the nodes of the graph that are the measurement points, and the links tha t are the
connections between measurement points.</t> connections between measurement points.</t>
<t>There are some techniques that can help with the building of the moni
<t>There are some techniques that can help with the building of the monit toring network (as an example,
oring network (as an example see <xref target="I-D.ietf-ippm-route" format="default"/>). In general, t
it is possible to mention <xref target="I-D.ietf-ippm-route"/>). In gener here are different options:
al there are different options: the monitoring network can be obtained by considering all the possible
the monitoring network can be obtained by considering all the possible paths paths for the traffic or periodically checking the traffic (e.g. daily,
for the traffic or also weekly, monthly) and updating the graph as appropriate,
by periodically checking the traffic (e.g. daily, weekly, monthly) and updat
e the graph as appropriate,
but this is up to the Network Management System (NMS) configuration.</t> but this is up to the Network Management System (NMS) configuration.</t>
<t>So a graph model of the monitoring network can be built according to
<t>So a graph model of the monitoring network can be built according to t the Alternate-Marking
he Alternate Marking
method: the monitored interfaces and links are identified. Only the measu rement points and links method: the monitored interfaces and links are identified. Only the measu rement points and links
where the traffic has flowed have to be represented in the graph.</t> where the traffic has flowed have to be represented in the graph.</t>
<t><xref target="monitored-graph"/> shows a simple example of a
<t>The following figure shows a simple example of a Monitoring Network gr monitoring network graph:</t>
aph:</t> <figure anchor="monitored-graph">
<name>Monitoring Network Graph</name>
<figure anchor="monitored-graph" title="Monitoring Network Graph"> <artwork name="" type="" align="left" alt=""><![CDATA[
<artwork><![CDATA[
+------+ +------+
<> R6 <>--- <> R6 <>---
/ +------+ / +------+
+------+ +------+ / +------+ +------+ /
<> R2 <>---<> R4 <> <> R2 <>---<> R4 <>
/ +------+ \ +------+ \ / +------+ \ +------+ \
/ \ \ +------+ / \ \ +------+
+------+ / +------+ \ +------+ <> R7 <>--- +------+ / +------+ \ +------+ <> R7 <>---
---<> R1 <>---<> R3 <>---<> R5 <> +------+ ---<> R1 <>---<> R3 <>---<> R5 <> +------+
+------+ \ +------+ \ +------+ \ +------+ \ +------+ \ +------+ \
skipping to change at line 521 skipping to change at line 359
\ \ <> R8 <>--- \ \ <> R8 <>---
\ \ +------+ \ \ +------+
\ \ \ \
\ \ +------+ \ \ +------+
\ <> R9 <>--- \ <> R9 <>---
\ +------+ \ +------+
\ \
\ +------+ \ +------+
<> R10 <>--- <> R10 <>---
+------+ +------+
]]></artwork> ]]></artwork>
</figure> </figure>
<t>Each monitoring point is characterized by the packet counter
<t>Each monitoring point is characterized by the packet counter
that refers only to a marking period of the monitored flow.</t> that refers only to a marking period of the monitored flow.</t>
<t>The same is also applicable for the delay, but it will be described
<t>The same is applicable also for the delay but it will be described
in the following sections.</t> in the following sections.</t>
</section>
</section> </section>
<section numbered="true" toc="default">
</section> <name>Multipoint Packet Loss</name>
<t>Since all the packets of the considered flow leaving the
<section title="Multipoint Packet Loss">
<t>Since all the packets of the considered flow leaving the
network have previously entered the network, the number of network have previously entered the network, the number of
packets counted by all the input nodes is always greater packets counted by all the input nodes is always greater than,
or equal than the number of packets counted by all the or equal to, the number of packets counted by all the
output nodes. Non-initial fragments are not considered here.</t> output nodes. Noninitial fragments are not considered here.</t>
<t>The assumption is the use of the Alternate-Marking method. In the case
<t>The assumption is the use of the Alternate Marking method. And of
in case of no packet loss occurring in the marking period, if all no packet loss occurring in the marking period, if all
the input and output points of the network domain to be monitored are the input and output points of the network domain to be monitored are
measurement points, the sum of the number of packets on all the measurement points, the sum of the number of packets on all the
ingress interfaces equals the number on egress interfaces for the ingress interfaces equals the number on egress interfaces for the
monitored flow. In this circumstance, if no packet loss occurs, monitored flow. In this circumstance, if no packet loss occurs,
the intermediate measurement points have only the task to split the intermediate measurement points only have the task of splitting
the measurement.</t> the measurement.</t>
<t>It is possible to define the Network Packet Loss of one monitored flow
<t>It is possible to define the Network Packet Loss of one monitored flow for a single period. In a packet network, the number of lost packets is t
for a single period: «In a packet network, the number of lost packets is he
the
number of packets counted by the input nodes minus the number of packets number of packets counted by the input nodes minus the number of packets
counted by the output nodes». This is true for every packet flow counted by the output nodes. This is true for every packet flow
in each marking period.</t> in each marking period.</t>
<t>The monitored network packet loss with n input nodes and m output nodes
<t>The Monitored Network Packet Loss with n input nodes and m output node
s
is given by:</t> is given by:</t>
<t>PL = (PI1 + PI2 +...+ PIn) - (PO1 + PO2 +...+ POm)</t>
<t>PL = (PI1 + PI2 +...+ PIn) - (PO1 + PO2 +...+ POm)</t> <t>where:</t>
<t>PL is the network packet loss (number of lost packets)</t>
<t>where:</t> <t>PIi is the number of packets flowed through the i-th input node in this
<t>PL is the Network Packet Loss (number of lost packets)</t> period</t>
<t>PIi is the Number of packets flowed through the i-th Input node in thi <t>POj is the number of packets flowed through the j-th output node in
s period</t> this period</t>
<t>POj is the Number of packets flowed through the j-th Output node in th <t>The equation is applied on a per-time-interval basis and a per-flow bas
is period</t> is:</t>
<ul empty="true" spacing="normal">
<t>The equation is applied on a per-time-interval basis and on an per-flow b <li>The reference interval is the Alternate-Marking period, as defined
asis:<list> in <xref target="RFC8321" format="default">RFC 8321</xref>.</li>
<li>The flow definition is generalized here. Indeed, as described before
<t>The reference interval is the Alternate Marking period as defined in < , a multipoint packet flow
xref target="RFC8321">RFC 8321</xref>.</t> is considered, and the identification fields can be selected without any
constraints.</li>
<t>The flow definition is generalized here, indeed, as described before, </ul>
a multipoint packet flow
is considered and the identification fields can be selected without any c
onstraints.</t>
</list></t>
</section> </section>
<section numbered="true" toc="default">
<section title="Network Clustering"> <name>Network Clustering</name>
<t>The previous Equation can determine the number of packets lost <t>The previous equation can determine the number of packets lost
globally in the monitored network, exploiting only the data provided globally in the monitored network, exploiting only the data provided
by the counters in the input and output nodes.</t> by the counters in the input and output nodes.</t>
<t>In addition, it is also possible to leverage the data provided by the o
<t>In addition it is also possible to leverage the data provided by the o ther counters
ther counters
in the network to converge on the smallest identifiable subnetworks where the losses occur. in the network to converge on the smallest identifiable subnetworks where the losses occur.
These subnetworks are named Clusters.</t> These subnetworks are named "clusters".</t>
<t>A cluster graph is a subnetwork of the entire monitoring network graph
<t>A Cluster graph is a subnetwork of the entire Monitoring Network graph that still
that still satisfies the packet loss equation (introduced in the previous section),
satisfies the packet loss equation (introduced in the previous section) w where PL in this case
here PL in this case is the number of packets lost in the cluster. As for the entire monitorin
is the number of packets lost in the Cluster. As for the entire Monitorin g network graph, the cluster
g Network graph, the Cluster
is defined on a per-flow basis.</t> is defined on a per-flow basis.</t>
<t>For this reason, a cluster should contain all the arcs emanating from i
<t>For this reason a Cluster should contain all the arcs emanating from i ts input nodes
ts input nodes
and all the arcs terminating at its output nodes. This ensures that we ca n and all the arcs terminating at its output nodes. This ensures that we ca n
count all the packets (and only those) exiting an input node count all the packets (and only those) exiting an input node
again at the output node, whatever path they follow.</t> again at the output node, whatever path they follow.</t>
<t>In a completely monitored unidirectional network (a network where every
<t>In a completely monitored unidirectional network (a network where ever
y
network interface is monitored), each network device corresponds network interface is monitored), each network device corresponds
to a Cluster and each physical link corresponds to two to a cluster, and each physical link corresponds to two
Clusters (one for each device).</t> clusters (one for each device).</t>
<t>Clusters can have different sizes depending on the flow-filtering crite
<t>Clusters can have different sizes depending on flow filtering criteria ria adopted.</t>
adopted.</t> <t>Moreover, sometimes clusters can be optionally simplified. For example,
when two monitored interfaces
<t>Moreover, sometimes Clusters can be optionally simplified. For example are divided by a single router (one is the input interface, the other is
when two monitored interfaces the output interface,
are divided by a single router (one is the input interface and the other
is the output interface
and the router has only these two interfaces), instead of counting exactl y twice, upon entering and leaving, and the router has only these two interfaces), instead of counting exactl y twice, upon entering and leaving,
it is possible to consider a single measurement point (in this case we do it is possible to consider a single measurement point. In this case,
not care of the internal packet loss we do not care about the internal packet loss of the router.</t>
of the router).</t>
<t>It is worth highlighting that it might also be convenient to define Cl
usters based on the topological
information and applicable to all the possible flows in the monitored net
work.</t>
<section title="Algorithm for Cluster partition">
<t>A simple algorithm can be applied in order to split our monitoring net
work into Clusters.
This can be done for each direction separately. The Cluster partition is
based on the Monitoring Network Graph
that can be valid for a specific flow or can also be general and valid fo
r the entire network topology.</t>
<t>It is a two-step algorithm:<list style="symbols">
<t>Group the links where there is the same starting node;</t>
<t>Join the grouped links with at least one ending node in common.</t>
</list></t>
<t>Considering that the links are unidirectional, the first step implies <t>It is worth highlighting that it might also be convenient to define clu
to list all the links as connection sters based on the topological
between two nodes and to group the different links if they have the same information so that they are applicable to all the possible flows in the
starting node. monitored network.</t>
Note that it is possible to start from any link and the procedure works <section numbered="true" toc="default">
anyway. Following this classification, <name>Algorithm for Clusters Partition</name>
the second step implies to eventually join the groups classified in the <t>A simple algorithm can be applied in order to split our monitoring ne
first step by looking at the ending nodes. twork into clusters.
This can be done for each direction separately. The clusters partition
is based on the monitoring network graph, which can be valid for a
specific flow or can also be general and valid for the entire network
topology.</t>
<t>It is a two-step algorithm:</t>
<ol spacing="normal">
<li>Group the links where there is the same starting node;</li>
<li>Join the grouped links with at least one ending node in common.</l
i>
</ol>
<t>Considering that the links are unidirectional, the first step
implies listing all the links as connections between two nodes and
grouping the different links if they have the same starting node.
Note that it is possible to start from any link, and the procedure
will work. Following this classification,
the second step implies eventually joining the groups classified in the
first step by looking at the ending nodes.
If different groups have at least one common ending node, they are put t ogether and belong to the same set. If different groups have at least one common ending node, they are put t ogether and belong to the same set.
After the application of the two steps of the algorithm, each one of the After the application of the two steps of the algorithm, each one of the
composed sets of links composed sets of links,
together with the endpoint nodes constitutes a Cluster.</t> together with the endpoint nodes, constitutes a cluster.</t>
<t>In our monitoring network graph example, it is possible to identify t
<t>In our monitoring network graph example it is possible to identify the he clusters partition
Clusters partition
by applying this two-step algorithm.</t> by applying this two-step algorithm.</t>
<t>The first step identifies the following groups:</t>
<t>The first step identifies the following groups:<list style="numbers"> <ol spacing="normal" type="1">
<t>Group 1: (R1-R2), (R1-R3), (R1-R10)</t> <li>Group 1: (R1-R2), (R1-R3), (R1-R10)</li>
<t>Group 2: (R2-R4), (R2-R5)</t> <li>Group 2: (R2-R4), (R2-R5)</li>
<t>Group 3: (R3-R5), (R3-R9)</t> <li>Group 3: (R3-R5), (R3-R9)</li>
<t>Group 4: (R4-R6), (R4-R7)</t> <li>Group 4: (R4-R6), (R4-R7)</li>
<t>Group 5: (R5-R8)</t> <li>Group 5: (R5-R8)</li>
</list></t> </ol>
<t>And then, the second step builds the clusters partition (in particula
<t>And then, the second step builds the Clusters partition (in particular r,
we can underline that Group 2 and Group 3 connect together, since R5 is i we can underline that Groups 2 and 3 connect together, since R5 is in
n common):</t>
common):<list style="numbers"> <ol spacing="normal" type="1">
<t>Cluster 1: (R1-R2), (R1-R3), (R1-R10)</t> <li>Cluster 1: (R1-R2), (R1-R3), (R1-R10)</li>
<t>Cluster 2: (R2-R4), (R2-R5), (R3-R5), (R3-R9)</t> <li>Cluster 2: (R2-R4), (R2-R5), (R3-R5), (R3-R9)</li>
<t>Cluster 3: (R4-R6), (R4-R7)</t> <li>Cluster 3: (R4-R6), (R4-R7)</li>
<t>Cluster 4: (R5-R8)</t> <li>Cluster 4: (R5-R8)</li>
</list></t> </ol>
<t>The flow direction here considered is from left to right. For the opp
<t>The flow direction here considered is from left to right. For the opposit osite direction,
e direction the same reasoning can be applied, and in this example, you get the
the same way of reasoning can be applied and, in this example, you get th same clusters partition.</t>
e same Clusters partition.</t> <t>In the end, the following 4 clusters are obtained:</t>
<figure anchor="clusters">
<t>In the end the following 4 Clusters are obtained:</t> <name>Clusters Example</name>
<artwork name="" type="" align="left" alt=""><![CDATA[
<figure anchor="clusters" title="Clusters example">
<artwork><![CDATA[
Cluster 1 Cluster 1
+------+ +------+
<> R2 <>--- <> R2 <>---
/ +------+ / +------+
/ /
+------+ / +------+ +------+ / +------+
---<> R1 <>---<> R3 <>--- ---<> R1 <>---<> R3 <>---
+------+ \ +------+ +------+ \ +------+
\ \
\ \
skipping to change at line 713 skipping to change at line 533
<> R7 <>--- <> R7 <>---
+------+ +------+
Cluster 4 Cluster 4
+------+ +------+
---<> R5 <> ---<> R5 <>
+------+ \ +------+ \
\ +------+ \ +------+
<> R8 <>--- <> R8 <>---
+------+ +------+
]]></artwork> ]]></artwork>
</figure> </figure>
<t>There are clusters with more than two nodes as well as two-node clust
<t>There are Clusters with more than 2 nodes and two-nodes Cluste ers.
rs. In the two-node clusters, the loss is on the link (Cluster 4).
In the two-nodes Clusters the loss is on the link (Cluster 4). In more-than-two-node clusters, the loss is on the cluster, but
In more-than-2-nodes Clusters the loss is on the Cluster but we cannot know in which link (Cluster 1, 2, or 3).</t>
we cannot know in which link (Cluster 1, 2, 3).</t> <t>In this way, the calculation of packet loss can be made on a cluster
basis.
<t>In this way the calculation of packet loss can be made on Clus Note that the packet counters for each marking period permit calc
ter basis. ulating the packet rate
Note that the packet counters for each marking period permit to c on a cluster basis, so Committed Information Rate (CIR) and Exces
alculate the packet rate s Information Rate (EIR)
on Cluster basis, so Committed Information Rate (CIR) and Excess could also be deduced on a cluster basis.</t>
Information Rate (EIR) <t>Obviously, by combining some clusters in a new connected subnetwork
could also be deduced on Cluster basis.</t> (called a "super cluster"), the packet-loss rule is still true.</
t>
<t>Obviously, by combining some Clusters in a new connected subne <t>In this way, in a very large network, there is no need to configure d
twork etailed
(called Super Cluster) the Packet Loss Rule is still true.</t>
<t>In this way, in a very large network there is no need to confi
gure detailed
filter criteria to inspect the traffic. You can check a multipoin t network and, filter criteria to inspect the traffic. You can check a multipoin t network and,
in case of problems, you can go deep with a step-by-step cluster analysis, in case of problems, go deep with a step-by-step cluster analysis ,
but only for the cluster or combination of clusters where the pro blem happens.</t> but only for the cluster or combination of clusters where the pro blem happens.</t>
<t>In summary, once defined a flow, the algorithm to build the Cl <t>In summary, once a flow is defined, the algorithm to build the cluste
uster Partition rs partition
considers all the possible links and nodes crossed by the given f is based on topological information; therefore, it considers all
low, even if the possible links and nodes crossed by the given flow, even if
there is no traffic. It is based on topological information. So, there is no traffic. So, if the flow does not
if the flow does not enter or traverse all the nodes, the counters have a nonzero
enter or traverse all the nodes, the counters have a non-zero val value for the involved nodes and a zero value for the other
ue for the involved nodes, nodes without traffic; but in the end, all the formulas
while a zero value for the other nodes without traffic, but, in t
he end all the formulas
are still valid.</t> are still valid.</t>
<t>The algorithm described above is an iterative clustering algorithm, b
ut it is also
possible to apply a recursive clustering algorithm by using the n
ode-node adjacency
matrix representation <xref target="IEEE-ACM-ToN-MPNPM" format="d
efault"/>.</t>
<t>The algorithm described above is an Iterative clustering algor <t>The complete and mathematical analysis of the possible algorithms for
ithm, but it is also clusters
possible to apply a Recursive clustering algorithm by using the n
ode-node adjacency
matrix representation (<xref target="IEEE-ACM-ToN-MPNPM"/>).</t>
<t>The complete and mathematical analysis of the possible Algorit
hms for Cluster
partition, including the considerations in terms of efficiency an d a comparison partition, including the considerations in terms of efficiency an d a comparison
between the different methods, is in the paper <xref target="IEEE between the different methods, is in the paper <xref
-ACM-ToN-MPNPM"/>.</t> target="IEEE-ACM-ToN-MPNPM" format="default"/>.</t>
</section>
</section>
</section> </section>
<section numbered="true" toc="default">
<section title="Timing Aspects"> <name>Timing Aspects</name>
<t>It is important to consider the timing aspects, since out-of-order pack
<t>It is important to consider the timing aspects, since out of order pa ets happen and have
ckets happen and have to be handled as well, as described in <xref target="RFC8321"
to be handled as well as described in <xref target="RFC8321">RFC 8321</x format="default">RFC 8321</xref>. However, in a
ref>. But, in a multisource situation, an additional issue has to be considered. With mu
multi-source situation an additional issue has to be considered. With mu ltipoint path,
ltipoint path,
the egress nodes will receive alternate marked packets in random order f rom different ingress nodes, the egress nodes will receive alternate marked packets in random order f rom different ingress nodes,
and this must not affect the measurement.</t> and this must not affect the measurement.</t>
<t>So, if we analyze a multipoint-to-multipoint path with more than one ma
<t>So, if we analyse a multipoint-to-multipoint path with more than one rking node,
marking node, it is important to recognize the reference measurement interval. In general
it is important to recognize the reference measurement interval. In general , the measurement
the measurement
interval for describing the results is the interval of the marking node that is more aligned with interval for describing the results is the interval of the marking node that is more aligned with
the start of the measurement, as reported in the following figure.</t> the start of the measurement, as reported in <xref target="measint" />.<
/t>
<t>Note that the mark switching approach based on a fixed timer is consi <t>Note that the mark switching approach based on a fixed timer is conside
dered in this document.</t> red in this document.</t>
<figure anchor="measint">
<figure anchor="measint" title="Measurement Interval"> <name>Measurement Interval</name>
<artwork><![CDATA[ <artwork name="" type="" align="left" alt=""><![CDATA[
time -> start stop time -> start stop
T(R1) |-------------| T(R1) |-------------|
T(R2) |-------------| T(R2) |-------------|
T(R3) |------------| T(R3) |------------|
]]></artwork> ]]></artwork>
</figure> </figure>
<t>In the figure it is assumed that the node with the earliest clock (R1)
identifies the right
starting and ending time of the measurement, but it is just an assumption
and other possibilities
could occur. So, in this case, T(R1) is the measurement interval and its
recognition is essential
in order to be compatible and make comparison with other active/passive/h
ybrid Packet Loss metrics.</t>
<t>When we expand to multipoint-to-multipoint flows, we have to consider
that
all source nodes mark the traffic and this adds more complexity.</t>
<t>Regarding the timing aspects of the methodology, <xref target="RFC8321 <t>In <xref target="measint"/>, it is assumed that the node with the
">RFC 8321</xref> already earliest clock (R1) identifies the right
starting and ending times of the measurement, but it is just an assumptio
n, and other possibilities
could occur. So, in this case, T(R1) is the measurement interval, and its
recognition is essential
in order to make comparisons with other active/passive/hybrid Packet Loss
metrics.</t>
<t>When we expand to multipoint-to-multipoint flows, we have to consider t
hat
all source nodes mark the traffic, and this adds more complexity.</t>
<t>Regarding the timing aspects of the methodology, <xref target="RFC8321"
format="default">RFC 8321</xref> already
describes two contributions that are taken into account: the clock error between network devices and describes two contributions that are taken into account: the clock error between network devices and
the network delay between measurement points.</t> the network delay between measurement points.</t>
<t>But we should now consider an additional contribution. Since all source
<t>But we should now consider an additional contribution. Since all sourc nodes mark the traffic,
e nodes mark the traffic, the source measurement intervals can be of different lengths and with dif
the source measurement intervals can be of different lengths and with dif ferent offsets, and this
ferent offsets and this mismatch m can be added to d, as shown in <xref target="mtiming"/>.</t>
mismatch m can be added to d, as shown in figure.</t> <figure anchor="mtiming">
<name>Timing Aspects for Multipoint Paths</name>
<figure anchor="mtiming" title="Timing Aspects for Multipoint paths"> <artwork name="" type="" align="left" alt=""><![CDATA[
<artwork><![CDATA[
...BBBBBBBBB | AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA | BBBBBBBBB... ...BBBBBBBBB | AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA | BBBBBBBBB...
|<======================================>| |<======================================>|
| L | | L |
...=========>|<==================><==================>|<==========... ...=========>|<==================><==================>|<==========...
| L/2 L/2 | | L/2 L/2 |
|<=><===>| |<===><=>| |<=><===>| |<===><=>|
m d | | d m m d | | d m
|<====================>| |<====================>|
available counting interval available counting interval
]]></artwork> ]]></artwork>
</figure> </figure>
<t>So the misalignment between the marking source routers gives an additio
<t>So the misalignment between the marking source routers gives an additi nal constraint,
onal constraint and the value of m is added to d (which already includes clock error and
and the value of m is added to d (that already includes clock error and n network delay).</t>
etwork delay).</t> <t>Thus, three different possible contributions are considered: clock erro
r between network devices,
<t>Thus, three different possible contributions are considered: clock err network delay between measurement points, and the misalignment between th
or between network devices, e marking source routers.</t>
network delay between measurement points and the misalignment between the <t>In the end, the condition that must be satisfied to enable the method t
marking source routers.</t> o function properly is that the
available counting interval must be &gt; 0, and that means:</t>
<t>In the end, the condition that must be satisfied to enable the method to <t>L - 2m - 2d &gt; 0.</t>
function properly is that the <t>This formula needs to be verified for each measurement point on the mul
available counting interval must be > 0, and that means:</t> tipoint path, where m is misalignment
between the marking source routers, while d, already introduced in <xref
<t>L - 2m - 2d > 0.</t> target="RFC8321" format="default">RFC 8321</xref>,
<t>This formula needs to be verified for each measurement point on the mu
ltipoint path, where m is misalignment
between the marking source routers, while d, already introduced in <xref
target="RFC8321">RFC 8321</xref>,
takes into account clock error and network delay between network nodes. T herefore, the mismatch between takes into account clock error and network delay between network nodes. T herefore, the mismatch between
measurement intervals must satisfy this condition.</t> measurement intervals must satisfy this condition.</t>
<t>Note that the timing considerations are valid for both packet loss and
<t>Note that the timing considerations are valid for both packet loss and delay measurements.</t>
delay measurements.</t>
</section> </section>
<section numbered="true" toc="default">
<section title="Multipoint Delay and Delay Variation"> <name>Multipoint Delay and Delay Variation</name>
<t>The same line of reasoning can be applied to Delay and Delay V <t>The same line of reasoning can be applied to delay and delay variation.
ariation. Similarly to the delay measurements defined in <xref target="RFC8321" form
Similarly to the delay measurements defined in <xref target="RFC8 at="default">RFC 8321</xref>,
321">RFC 8321</xref>, the marking batches anchor the samples to a particular period, and this is
the marking batches anchor the samples to a particular period and the
this is the time reference that can be used.
time reference that can be used. It is important to highlight that both delay and delay-variation measureme
It is important to highlight that both delay and delay variation nts
measurements make sense in a multipoint path. The delay variation is calculated by cons
make sense in a multipoint path. The Delay Variation is calculate idering
d by considering the same packets selected for measuring the delay.</t>
the same packets selected for measuring the Delay.</t> <t>In general, it is possible to perform delay and delay-variation measure
ments
<t>In general, it is possible to perform delay and delay variatio on the basis of multipoint paths or single packets:</t>
n measurements <ul spacing="normal">
on multipoint paths basis or on single packets basis:<list style= <li>Delay measurements on the basis of multipoint paths mean that the de
"symbols"> lay value is representative
of an entire multipoint path (e.g., the whole multipoint network, a clust
<t>Delay measurements on multipoint paths basis means that the de er, or a combination
lay value is representative of clusters).</li>
of an entire multipoint path (e.g. whole multipoint network, a cl <li>Delay measurements on a single-packet basis mean that you can use
uster or a combination a multipoint path just
of clusters).</t> to easily couple packets between input and output nodes of a multipoint p
ath, as
<t>Delay measurements on a single packet basis means that you can described in the following sections.</li>
use multipoint path just </ul>
to easily couple packets between input and output nodes of a mult <section numbered="true" toc="default">
ipoint path, as it is <name>Delay Measurements on a Multipoint-Paths Basis</name>
described in the following sections.</t> <section numbered="true" toc="default">
</list></t> <name>Single-Marking Measurement</name>
<t>Mean delay and mean delay-variation measurements can also be genera
<section title="Delay measurements on multipoint paths basis"> lized
<section title="Single Marking measurement">
<t>Mean delay and mean delay variation measurements can also be g
eneralized
to the case of multipoint flows. It is possible to compute the av erage to the case of multipoint flows. It is possible to compute the av erage
one-way delay of packets, in one block, in a cluster or in the en tire one-way delay of packets in one block, a cluster, or the entire
monitored network.</t> monitored network.</t>
<t>The average latency can be measured as the difference
<t>The average latency can be measured as the difference
between the weighted averages of the mean timestamps of the between the weighted averages of the mean timestamps of the
sets of output and input nodes. This means that, in the calculati on, sets of output and input nodes. This means that, in the calculati on,
it is possible to weigh the timestamps by considering the number of packets it is possible to weigh the timestamps by considering the number of packets
for each endpoints.</t> for each endpoints.</t>
</section> </section>
</section> </section>
<section numbered="true" toc="default">
<section title="Delay measurements on single packets basis"> <name>Delay Measurements on a Single-Packet Basis</name>
<section title="Single and Double Marking measurement"> <section numbered="true" toc="default">
<t>Delay and delay variation measurements relative to only one pi <name>Single- and Double-Marking Measurement</name>
cked packet per period (both single <t>Delay and delay-variation measurements relative to only one picked
and double marked) can be performed in the Multipoint scenario wi packet per period (both single
th some limitations:<list style="hanging"> and double marked) can be performed in the multipoint scenario, w
<t>Single marking based on the first/last packet of the interval ith some limitations:</t>
would not work, because
it would not be possible to agree on the first packet of the inte
rval.</t>
<t>Double marking or multiplexed marking would work, but each mea <ul empty="true" spacing="normal">
surement would only <li>Single marking based on the first/last packet of the interval wo
uld not work, because
it would not be possible to agree on the first packet of the inte
rval.</li>
<li>Double marking or multiplexed marking would work, but each measu
rement would only
give information about the delay of a single path. However, by re peating the measurement give information about the delay of a single path. However, by re peating the measurement
multiple times, it is possible to get information about all the p aths in the multipoint flow. multiple times, it is possible to get information about all the p aths in the multipoint flow.
This can be done in case of point-to-multipoint path but it is mo This can be done in the case of a point-to-multipoint path, but
re difficult to achieve in case of it is more difficult to achieve in the case of a
multipoint-to-multipoint path because of the multiple source rout multipoint-to-multipoint path because of the multiple source rout
ers.</t> ers.</li>
</list></t> </ul>
<t>If we would perform a delay measurement for more than one
<t>If we would perform a delay measurement for more than one pick picked packet in the same marking period, and especially if
ed packet in the same marking period we want to get delay measurements on a
and, especially, if we want to get delay measurements on multipoi multipoint-to-multipoint basis, neither the single- nor the
nt-to-multipoint basis, double-marking method is useful in the multipoint scenario,
both single and double marking method are not useful in the Multi since they would not be representative of the entire
point scenario, since flow. The packets can follow different paths with various
they would not be representative of the entire flow. The packets delays, and in general it can be very difficult to recognize
can follow different paths marked packets in a multipoint-to-multipoint path,
with various delays, and in general it can be very difficult to r especially in the case when there is more than one per
ecognize marked packets period.</t>
in a multipoint-to-multipoint path especially in the case when th <t>A desirable option is to monitor simultaneously all the paths of a
ere is more than one per period.</t> multipoint path
in the same marking period; for this purpose, hashing can be used
<t>A desirable option is to monitor simultaneously all the paths , as reported in the next section.</t>
of a multipoint path </section>
in the same marking period and, for this purpose, hashing can be <section numbered="true" toc="default">
used as reported in the next Section.</t> <name>Hashing Selection Method</name>
<t>RFCs <xref target="RFC5474" sectionFormat="bare">5474</xref> and
</section> <xref target="RFC5475" sectionFormat="bare">5475</xref>
introduce sampling and filtering techniques for IP packet selecti
<section title="Hashing selection method"> on.</t>
<t><xref target="RFC5474">RFC 5474</xref> and <xref target="RFC54 <t>The hash-based selection methodologies for delay measurement can wo
75">RFC 5475</xref> rk in a
introduce sampling and filtering techniques for IP Packet Selecti multipoint-to-multipoint path and can be used either coupled
on.</t> to mean delay or stand-alone.</t>
<t><xref target="I-D.mizrahi-ippm-compact-alternate-marking" format="d
<t>The hash-based selection methodologies for delay measurement c efault"/> introduces how
an work in a to use the hash method (RFCs <xref target="RFC5474"
multipoint-to-multipoint path and can be used both coupled to mea sectionFormat="bare">5474</xref> and <xref target="RFC5475"
n delay or stand alone.</t> sectionFormat="bare">5475</xref>)
combined with the Alternate-Marking method for point-to-point flo
<t><xref target="I-D.mizrahi-ippm-compact-alternate-marking"/> introd ws.
uces how It is also called Mixed Hashed Marking: the coupling of a marking method
to use the Hash method (<xref target="RFC5474">RFC 5474</xref> an and hashing
d <xref target="RFC5475">RFC 5475</xref>) technique is very useful, because the marking batches anchor the
combined with Alternate Marking method for point-to-point flows. samples selected
It is also called Mixed Hashed Marking: the coupling of marking method a with hashing, and this simplifies the correlation of the hashing
nd hashing packets along the path.</t>
technique is very useful because the marking batches anchor the s <t>It is possible to use a basic-hash or a dynamic-hash method.
amples selected One of the challenges of the basic approach is that the frequency
with hashing and this simplifies the correlation of the hashing p of the sampled packets may vary considerably.
ackets along the path.</t>
<t>It is possible to use a basic hash or a dynamic hash method. For this
One of the challenges of the basic approach is that the frequency reason, the dynamic approach has been introduced for
of point-to-point flows in order to have the desired and
the sampled packets may vary considerably. For this reason the dy almost fixed number of samples for each measurement
namic period. Using the hash-based sampling, the number of
approach has been introduced for point-to-point flow in order to samples may vary a lot because it depends on the
have packet rate that is variable. The dynamic
the desired and almost fixed number of samples for each measureme approach helps to have an almost fixed number of
nt period. samples for each marking period, and this is a better
option for making regular measurements over time.
In the hash-based sampling, Alternate Marking is used to In the hash-based sampling, Alternate Marking is used to
create periods, so that hash-based samples are divided into batch create periods, so that hash-based samples are divided into batch
es, es, which
allowing to anchor the selected samples to their period. Moreove allows anchoring the selected samples to their period.
r in the dynamic Moreover, in the dynamic
hash-based sampling, by dynamically adapting the length of the ha sh value, hash-based sampling, by dynamically adapting the length of the ha sh value,
the number of samples is bounded in each marking period. the number of samples is bounded in each marking period.
This can be realized by choosing the maximum number of samples This can be realized by choosing the maximum number of samples
(NMAX) to be caught in a marking period. The algorithm starts (NMAX) to be caught in a marking period. The algorithm starts
with only few hash bits, that permit to select a greater percenta with only a few hash bits, which permits selecting a greater perc
ge entage
of packets (e.g. with 0 bit of hash all the packets are sampled, of packets (e.g., with 0 bits of hash all the packets are sampled
,
with 1 bit of hash half of the packets are sampled, and so on). with 1 bit of hash half of the packets are sampled, and so on).
When the number of selected packets reaches NMAX, a hashing bit i s When the number of selected packets reaches NMAX, a hashing bit i s
added. As a consequence, the sampling proceeds at half of the added. As a consequence, the sampling proceeds at half of the
original rate and also the packets already selected that do not m atch original rate, and also the packets already selected that do not match
the new hash are discarded. This step can be repeated iterativel y. the new hash are discarded. This step can be repeated iterativel y.
It is assumed that each sample includes the timestamp (used for d elay measurement) and It is assumed that each sample includes the timestamp (used for d elay measurement) and
the hash value, allowing the management system to match the sampl es received the hash value, allowing the management system to match the sampl es received
from the two measurement points. from the two measurement points.
The dynamic process statistically converges at the end of a marki ng The dynamic process statistically converges at the end of a marki ng
period and the final number of selected samples is between NMAX/2 and NMAX. period, and the final number of selected samples is between NMAX/ 2 and NMAX.
Therefore, the dynamic approach paces the sampling rate, allowing to bound the Therefore, the dynamic approach paces the sampling rate, allowing to bound the
number of sampled packets per sampling period.</t> number of sampled packets per sampling period.</t>
<t>In a multipoint environment, the behavior is similar to a point-to-
<t>In a multipoint environment the behaviour is similar to a poin point
t-to point
flow. In particular, in the context of a multipoint-to-multipoint flow, the flow. In particular, in the context of a multipoint-to-multipoint flow, the
dynamic hash could be the solution to perform delay measurements dynamic hash could be the solution for performing delay
on specific packets measurements on specific packets
and to overcome the single and double marking limitations.</t> and overcoming the single- and double-marking limitations.</t>
<t>The management system receives the samples, including the timestamp
<t>The management system receives the samples including the times s and
tamps and the hash value, from all the MPs, and this happens for both point
the hash value from all the MPs, and this happens both for point- -to-point
to-point and multipoint-to-multipoint flows. Then, the longest hash used
and for multipoint-to-multipoint flows. Then the longest hash use by the MPs is deduced and applied to couple timestamps from eithe
d by MPs is deduced r the same packets of
and it is applied to couple timestamps of the same packets of 2 M 2 MPs of a point-to-point path, or the input and output MPs of a
Ps of a point-to-point path cluster (or a super cluster or the entire network).
or of input and output MPs of a Cluster (or a Super Cluster or th But some considerations are needed: if there isn't packet loss, t
e entire network). he set of input samples
But some considerations are needed: if there isn't packet loss th is always equal to the set of output samples. In the case of pac
e set of input samples ket loss, the set of
is always equal to the set of output samples. In case of packet output samples can be a subset of input samples, but the
loss the set of method still works because, at the end,
output samples can be a subset of input samples but the method st
ill works because, at the end,
it is easy to couple the input and output timestamps of each caug ht packet using the hash it is easy to couple the input and output timestamps of each caug ht packet using the hash
(in particular the “unused part of the hash” that should be diffe (in particular, the "unused part of the hash" that should be diff
rent for each packet).</t> erent for each packet).</t>
<t>Therefore, the basic hash is logically similar to the double-markin
<t>Therefore, the basic hash is logically similar to the double m g method, and
arking method, and in the case of a point-to-point path, double-marking and basic-ha
in case of point-to-point path double marking and basic hash sele sh selection are equivalent.
ction are equivalent. The dynamic approach scales the number of measurements per
The dynamic approach scales the number of measurements per interv interval. It would seem
al, and it would seem
that double marking would also work well if we reduced the interv al length, but that double marking would also work well if we reduced the interv al length, but
this can be done only for point-to-point path and not for multipo this can be done only for a point-to-point path and not for a mul
int path, where we tipoint path, where we
cannot couple the picked packets in a multipoint paths. cannot couple the picked packets in a multipoint path.
So, in general, if we want to get delay measurements on multipoin So, in general, if we want to get delay measurements on the
t-to-multipoint path basis of a multipoint-to-multipoint path, and want to select more
basis and want to select more than one packet per period, double than one packet per period, double marking cannot be used
marking cannot be used
because we could not be able to couple the picked packets between input and output nodes. because we could not be able to couple the picked packets between input and output nodes.
On the other hand we can do that by using hashing selection.</t> On the other hand, we can do that by using hashing selection.</t>
</section>
</section> </section>
</section>
</section> <section numbered="true" toc="default">
<name>A Closed-Loop Performance-Management Approach</name>
</section> <t>The Multipoint Alternate-Marking framework that is introduced in this d
ocument adds flexibility
<section title="A Closed Loop Performance Management approach"> to Performance Management (PM), because it can reduce the order of magnit
ude of the packet counters.
<t>The Multipoint Alternate Marking framework that is introduced in this This allows an SDN orchestrator to supervise, control, and manage PM in l
document adds flexibility arge networks.</t>
to Performance Management (PM) because it can reduce the order of magnitu
de of the packet counters.
This allows an SDN Orchestrator to supervise, control and manage PM in la
rge networks.</t>
<t>The monitoring network can be considered as a whole or can be split in
Clusters, that are the
smallest subnetworks (group-to-group segments), maintaining the packet lo
ss property for each subnetwork.
They can also be combined in new connected subnetworks at different level
s depending on the detail we want
to achieve.</t>
<t>An SDN Controller or a Network Management System (NMS) can calibrate P <t>The monitoring network can be considered as a whole or split into clust
erformance Measurements ers that are the
smallest subnetworks (group-to-group segments), maintaining the
packet-loss property for each subnetwork. The clusters can also be
combined in new, connected subnetworks at different levels, depending on
the detail we want to achieve.</t>
<t>An SDN controller or a Network Management System (NMS) can calibrate pe
rformance measurements,
since they are aware of the network topology. They can start without exam ining in depth. In case of necessity since they are aware of the network topology. They can start without exam ining in depth. In case of necessity
(packet loss is measured or the delay is too high), the filtering criteri a could be immediately reconfigured (packet loss is measured or the delay is too high), the filtering criteri a could be immediately reconfigured
in order to perform a partition of the network by using Clusters and/or d in order to perform a partition of the network by using clusters and/or d
ifferent combinations of Clusters. ifferent combinations of clusters.
In this way the problem can be localized in a specific Cluster or in a si In this way, the problem can be localized in a specific cluster or a sing
ngle combination of Clusters le combination of clusters,
and a more detailed analysis can be performed step-by-step by successive and a more detailed analysis can be performed step by step by successive
approximation up to approximation up to
a point-to-point flow detailed analysis. This is the so called Closed Loo a point-to-point flow detailed analysis. This is the so-called "closed lo
p.</t> op".</t>
<t>This approach can be called "network zooming" and can be performed in t
<t>This approach can be called Network Zooming and can be performed in tw wo different ways:</t>
o different ways:</t> <t>1) change the traffic filter and select more detailed flows;</t>
<t>1) change the traffic filter and select more detailed flows;</t> <t>2) activate new measurement points by defining more specified clusters.
<t>2) activate new measurement points by defining more specified clusters </t>
.</t> <t>The network-zooming approach implies that some filters or rules are cha
nged and that therefore there is a
<t>The Network Zooming approach implies that the some filters or rules ar transient time to wait once the new network configuration takes effect. T
e changed and there is a his time can be determined
transient time to wait once the new network configuration takes effect an
d it can be determined
by the Network Orchestrator/Controller, based on the network conditions.< /t> by the Network Orchestrator/Controller, based on the network conditions.< /t>
<t>For example, if the Network Zooming identifies the performance problem for the traffic coming from <t>For example, if the network zooming identifies the performance problem for the traffic coming from
a specific source, we need to recognize the marked signal from this speci fic source node and its relative path. a specific source, we need to recognize the marked signal from this speci fic source node and its relative path.
For this purpose we can activate all the available measurement points and For this purpose, we can activate all the available measurement points
specify better the flow filter criteria and better specify the flow filter criteria
(i.e. 5-tuple). As an alternative, it can be enough to select packets fro (i.e., 5-tuple). As an alternative, it can be enough to select packets
m the specific source for delay measurements, from the specific source for delay measurements;
and in this case it is possible to apply the hashing technique as mention in this case, it is possible to apply the hashing technique, as mentioned
ed in the previous sections.</t> in the previous sections.</t>
<t><xref target="I-D.song-opsawg-ifit-framework" format="default"/> define
<t><xref target="I-D.song-opsawg-ifit-framework"/> defines an architectur s an architecture where the centralized
e where the centralized Data Collector and Network Management can apply the intelligent and flexi
Data Collector and Network Management can apply the intelligent and flexi ble Alternate-Marking algorithm
ble Alternate Marking algorithm
as previously described.</t> as previously described.</t>
<t>As for <xref target="RFC8321" format="default">RFC 8321</xref>, it is p
<t>As for <xref target="RFC8321">RFC 8321</xref>, it is possible to class ossible to classify the traffic and mark a portion of
ify the traffic and mark a portion of the total traffic. For each period, the packet rate and bandwidth are cal
the total traffic. For each period the packet rate and bandwidth are calc culated from the number of packets.
ulated from the number of packets. In this way, the network orchestrator becomes aware if the traffic rate s
In this way the Network Orchestrator becomes aware if the traffic rate ov urpasses limits.
ercomes limits. In addition, more precision can be obtained by reducing the marking perio
In addition more precision can be obtained by reducing the marking period d; indeed, some implementations use
, indeed some implementations use a marking period of 1 sec or less.</t>
a marking period of 1 sec and less.</t> <t>In addition, an SDN controller could also collect the measurement histo
ry.</t>
<t>In addition an SDN Controller could also collect the measurement histo <t>It is important to mention that the Multipoint Alternate Marking framew
ry.</t> ork also helps Traffic Visualization.
Indeed, this methodology is very useful for identifying which path or clu
<t>It is important to mention that the Multipoint Alternate Marking framewor ster is crossed by the flow.</t>
k also helps Traffic Visualization. </section>
Indeed this methodology is very useful to identify which path or which cl <section numbered="true" toc="default">
uster is crossed by the flow.</t> <name>Examples of Application</name>
<t>There are application fields where it may be useful to take into
</section>
<section title="Examples of application">
<t>There are application fields where it may be useful to take into
consideration the Multipoint Alternate Marking:</t> consideration the Multipoint Alternate Marking:</t>
<dl>
<t><list style="symbols"> <dt>VPN:</dt><dd>The IP traffic is selected on an IP-source basis in bot
<t>VPN: The IP traffic is selected on IP source basis in both h
directions. At the endpoint WAN interface all the output traffic directions. At the endpoint WAN interface, all the output traffic
is counted in a single flow. The input traffic is composed by all is counted in a single flow. The input traffic is composed of all
the other flows aggregated for source address. So, by considering the other flows aggregated for source address. So, by considering
n end-points, the monitored flows are n (each flow with 1 ingress n endpoints, the monitored flows are n (each flow with 1 ingress
point and (n-1) egress points) instead of n*(n-1) flows (each point and (n-1) egress points) instead of n*(n-1) flows (each
flow, with 1 ingress point and 1 egress point);</t> flow, with 1 ingress point and 1 egress point).</dd>
<t>Mobile Backhaul: LTE traffic is selected, in the Up direction, by <dt>Mobile Backhaul:</dt><dd>LTE traffic is selected, in the Up directio
the EnodeB source address and, in Down direction, by the EnodeB n, by
destination address because the packets are sent from the Mobile the EnodeB source address and, in the Down direction, by the EnodeB
Packet Core to the EnodeB. So the monitored flow is only one per destination address, because the packets are sent from the Mobile
EnodeB in both directions;</t> Packet Core to the EnodeB. So the monitored flow is only one per
<t>Over The Top (OTT) services: The traffic is selected, in the Down dire EnodeB in both directions.</dd>
ction <dt>Over The Top (OTT) services:</dt><dd>The traffic is selected, in the
by the source addresses of the packets sent by OTT Servers. In Down direction,
the opposite direction (Up) by the destination IP addresses of the by the source addresses of the packets sent by OTT servers. In
same Servers. So the monitoring is based on a single flow per OTT the opposite direction (Up), it is selected by the destination IP addresse
Servers in both directions.</t> s of the
<t>Enterprise SD-WAN: SD-WAN allows to connect remote branch offices to same servers. So the monitoring is based on a single flow per OTT
Data Centers and build higher-performance WANs. A centralized controlle server in both directions.</dd>
r <dt>Enterprise SD-WAN:</dt><dd>SD-WAN allows connecting remote branch of
fices to
data centers and building higher-performance WANs. A centralized contro
ller
is used to set policies and prioritize traffic. The SD-WAN takes into a ccount is used to set policies and prioritize traffic. The SD-WAN takes into a ccount
these policies and the availability of network bandwidth to route traff ic. these policies and the availability of network bandwidth to route traff ic.
This helps ensure that application performance meets service level agre ements (SLAs). This helps ensure that application performance meets Service Level Agre ements (SLAs).
This methodology can also help the path selection for the WAN connectio n based on This methodology can also help the path selection for the WAN connectio n based on
per Cluster and per flow performance. per-cluster and per-flow performance.
</t> </dd>
</list></t> </dl>
<t>Note that the preceding list is just an example and is not exhaustive.
<t>Note that the list is just an example and it is not exhaustive. More a More applications
pplications
are possible.</t> are possible.</t>
</section>
<section title="Security Considerations">
<t>This document specifies a method to perform measurements that
does not
directly affect Internet security nor applications that run on th
e Internet.
However, implementation of this method must be mindful of securit
y
and privacy concerns, as explained in <xref target="RFC8321">RFC
8321</xref>.</t>
</section> </section>
<section numbered="true" toc="default">
<section anchor="Acknowledgements" title="Acknowledgements"> <name>Security Considerations</name>
<t>The authors would like to thank Al Morton, Tal Mizrahi, Rachel Huang <t>This document specifies a method of performing measurements that does n
for the precious ot
contribution.</t> directly affect Internet security or applications that run on the Internet
.
However, implementation of this method must be mindful of security
and privacy concerns, as explained in <xref target="RFC8321" format="defau
lt">RFC 8321</xref>.</t>
</section> </section>
<!-- Possibly a 'Contributors' section ... --> <section anchor="IANA" numbered="true" toc="default">
<name>IANA Considerations</name>
<section anchor="IANA" title="IANA Considerations"> <t>This document has no IANA actions.</t>
<t>This memo makes no requests of IANA.</t>
</section> </section>
</middle>
</middle>
<!-- *****BACK MATTER ***** -->
<back> <back>
<!-- References split to informative and normative -->
<references title="Normative References">
<?rfc include='reference.RFC.5644'?> <displayreference target="I-D.mizrahi-ippm-compact-alternate-marking" to="ALTERN
ATE-MARKING"/>
<displayreference target="I-D.zhou-ippm-enhanced-alternate-marking" to="ENHANCED
-ALTERNATE-MARKING"/>
<displayreference target="I-D.song-opsawg-ifit-framework" to="IFIT-FRAMEWORK"/>
<displayreference target="I-D.ietf-ippm-route" to="ROUTE-ASSESSMENT"/>
<?rfc include='reference.RFC.8321'?> <references>
<name>References</name>
<references>
<name>Normative References</name>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://xml
2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5644.xml"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://xml
2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8321.xml"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://xml
2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5474.xml"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://xml
2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5475.xml"/>
</references>
<references>
<name>Informative References</name>
<?rfc include='reference.RFC.5474'?> <reference anchor="IEEE-ACM-ToN-MPNPM">
<front>
<title>Multipoint Passive Monitoring in Packet Networks</title>
<author surname="Cociglio" initials="M">
<organization />
</author>
<author surname="Fioccola" initials="G">
<organization />
</author>
<author surname="Marchetto" initials="G">
<organization />
</author>
<author surname="Sapio" initials="A">
<organization />
</author>
<author surname="Sisto" initials="R">
<organization />
</author>
<date year="2019" month="December"/>
</front>
<seriesInfo name="IEEE/ACM Transactions on Networking"
value="vol. 27, no. 6"/>
<seriesInfo name="pp." value="2377-2390"/>
<seriesInfo name="DOI" value="10.1109/TNET.2019.2950157"/>
</reference>
<?rfc include='reference.RFC.5475'?> <reference anchor="IEEE-Network-PNPM">
<front>
<title>AM-PM: Efficient Network Telemetry using Alternate Marking</t
itle>
<author surname="Mizrahi" initials="T">
<organization/>
</author>
<author surname="Navon" initials="G">
<organization/>
</author>
<author surname="Fioccola" initials="G">
<organization/>
</author>
<author surname="Cociglio" initials="M">
<organization/>
</author>
<author surname="Chen" initials="M">
<organization/>
</author>
<author surname="Mirsky" initials="G">
<organization/>
</author>
<date year="2019" month="July"/>
</front>
<seriesInfo name="IEEE Network" value="vol. 33, no. 4"/>
<seriesInfo name="pp." value="155-161"/>
<seriesInfo name="DOI" value="10.1109/MNET.2019.1800152"/>
</reference>
</references> <!-- [rfced] [I-D.mizrahi-ippm-compact-alternate-marking] IESG state Expired - ->
<references title="Informative References"> <xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://www
<!-- A reference written by by an organization not a persoN. --> .rfc-editor.org/refs/bibxml3/reference.I-D.mizrahi-ippm-compact-alternate-markin
<reference anchor='IEEE-ACM-ToN-MPNPM'> g.xml"/>
<front>
<title>Multipoint Passive Monitoring in Packet Networks</title>
<author>
<organization>IEEE/ACM TRANSACTION ON NETWORKING</organization>
</author>
<date year='2019' />
</front>
<seriesInfo name='DOI' value='10.1109/TNET.2019.2950157'/>
</reference>
<reference anchor='IEEE-Network-PNPM'> <!-- [rfced] [I-D.zhou-ippm-enhanced-alternate-marking] IESG state I-D Exists --
<front> >
<title>AM-PM: Efficient Network Telemetry using Alternate Marking</title>
<author>
<organization>IEEE Network</organization>
</author>
<date year='2019' />
</front>
<seriesInfo name='DOI' value='10.1109/MNET.2019.1800152'/>
</reference>
<?rfc include='reference.I-D.mizrahi-ippm-compact-alternate-marking'?> <xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://www .rfc-editor.org/refs/bibxml3/reference.I-D.zhou-ippm-enhanced-alternate-marking. xml" />
<?rfc include='reference.I-D.zhou-ippm-enhanced-alternate-marking'?> <!-- [rfced] [I-D.song-opsawg-ifit-framework] IESG state I-D Exists -->
<?rfc include='reference.I-D.song-opsawg-ifit-framework'?> <xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://www .rfc-editor.org/refs/bibxml3/reference.I-D.song-opsawg-ifit-framework.xml"/>
<?rfc include='reference.I-D.ietf-ippm-route'?> <!-- [rfced] [I-D.ietf-ippm-route] IESG state Approved-announcement to be sent - ->
<?rfc include='reference.RFC.7011'?> <xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://www .rfc-editor.org/refs/bibxml3/reference.I-D.ietf-ippm-route.xml"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude" href="https://xml
2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7011.xml"/>
</references>
</references> </references>
<section anchor="Acknowledgements" numbered="false" toc="default">
</back> <name>Acknowledgements</name>
<t>The authors would like to thank <contact fullname="Al Morton" />,
<contact fullname="Tal Mizrahi"/>, and <contact fullname="Rachel Huang"/>
for the precious contributions.</t>
</section>
</back>
</rfc> </rfc>
 End of changes. 123 change blocks. 
1106 lines changed or deleted 857 lines changed or added

This html diff was produced by rfcdiff 1.48. The latest version is available from http://tools.ietf.org/tools/rfcdiff/