Performance Measurement (PM) with Alternate Marking Method in Service Function Chaining (SFC) DomainZTE Corp.gregimirsky@gmail.comTelecom Italiagiuseppe.fioccola@telecomitalia.itMarvell6 Hamada St.YokneamIsraeltalmi@marvell.com
Routing
SFC Working GroupInternet-DraftSFCOAMPerformance Measurement
This document describes how the alternate marking method be used
as the passive performance measurement method
in a Service Function Chaining (SFC) domain.
introduced architecture of a Service Function
Chain (SFC) in the network and defined its components as classifier,
Service Function Forwarder (SFF), and Service Function (SF).
describes passive performance measurement method,
which can be used to measure packet loss,
latency and jitter on live traffic. Because this method is based on marking consecutive batches of
packets the method often referred as Alternate Marking Method (AMM).
This document defines how the alternate marking method can be used to measure packet loss and delay metrics
of a service flow over e2e or any segment of the SFC.
MM: Marking Method OAM: Operations, Administration and MaintenanceSFC: Service Function ChainSF: Service FunctionSFF: Service Function ForwarderSFP: Service Function PathNSH: Network Service Header
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED",
"MAY", and "OPTIONAL" in this document are to be interpreted as
described in BCP 14
when, and only when, they appear in all capitals, as shown here.
defines format of the Network Service Header (NSH).
This document defines two bit long field, referred as Mark field (M in ,
as part of NSH Base and designated
for the alternate marking performance measurement method .
The Mark field MUST NOT be used in defining forwarding
and/or quality of service treatment of a SFC packet. The Mark field MUST be used only for the
performance measurement of data traffic in SFC layer. Because setting of the field to any value does not affect
forwarding and/or quality of service treatment of a packet, the alternate marking method in SFC layer can be viewed as true
example of passive performance measurement method.
The displays format of the Mark field.
where:
L- Loss flag;
D - Delay flag.
The marking method can be successfully used in the SFC.
Without limiting any generality consider SFC presented in .
Any combination of markings, Loss and/or Delay, can be applied to a service flow by
any component of the SFC at either ingress or egress point
to perform node, link, segment or end-to-end measurement to detect
performance degradation defect and localize it efficiently.
Using the marking method a component of the SFC creates distinct sub-flows in
the particular service traffic over SFC. Each sub-flow consists of consecutive
blocks that are unambiguously recognizable by a monitoring point at any component
of the SFC and can be measured to calculate packet loss and/or packet delay metrics.
As explained in the , marking can be
applied to delineate blocks of packets
based either on equal number of packets in a block or based on equal time interval.
The latter method offers better control
as it allows better account for capabilities of downstream nodes to report
statistics related to batches of packets and, at the same time,
time resolution that affects defect detection interval.
If the Single Mark measurement used, then the Delay flag MUST be set to zero on
transmit and ignored on reception by monitoring point.
The Loss flag is used to create alternate flows to measure the packet loss by
switching value of the Loss flag every N-th packet or at certain time intervals.
Delay metrics MAY be calculated with the alternate flow using any of the
following methods:
First/Last Packet Delay calculation: whenever the marking, i.e. value
of Loss flag, changes a component of the SFC can store the timestamp of the first/last
packet of the block. The timestamp can be compared with the timestamp of the
packet that arrived in the same order through a monitoring
point at downstream component of the SFC to compute packet delay. Because
timestamps collected based on order of arrival this method is sensitive to packet
loss and re-ordering of packets
Average Packet Delay calculation: an average delay is calculated by
considering the average arrival time of the packets within a single block.
A component of the SFC may collect timestamps for each packet received within
a single block. Average of the timestamp is the sum of all the timestamps
divided by the total number of packets received. Then difference between
averages calculated at two monitoring points is the average packet delay
on that segment. This method is robust to out of order packets and also
to packet loss (only a small error is introduced). This method only provides
single metric for the duration of the block and it doesn't give the minimum
and maximum delay values. This limitation could be overcome by reducing the
duration of the block by means of an
highly optimized implementation of the method.
Double Mark method allows measurement of minimum and maximum delays for the monitored flow
but it requires more nodal and network resources. If the Double Mark method used, then
the Loss flag MUST be used to create the alternate flow, i.e. mark larger batches of packets. The Delay flag
MUST be used to mark single packets to measure delay jitter.
The first marking (Loss flag alternation) is needed for packet loss and
also for average delay measurement. The second marking (Delay flag is put
to one) creates a new set of marked packets that are fully identified
over the SFC, so that a component can store the timestamps of these
packets; these timestamps can be compared with the timestamps of the
same packets on another component of the SFC to compute packet delay
values for each packet. The number of measurements can be easily increased by
changing the frequency of the second marking. But the frequency of the
second marking must be not too high in order to avoid out of order
issues. This method is useful to have not only the average delay but
also the minimum and maximum delay values and, in wider terms, to know
more about the statistic distribution of delay values.
Residence time is the variable part of the propagation delay that a packet experiences traversing a network, e.g. SFC.
Residence Time over an SFC is the sum of the nodal residence times, i.e. periods that the packet spent in each of SFFs that
compose the SFC. The nodal residence time in SFC itself is the sum of sub-nodal residence times that the packet spent in each of SFs that are
part of the given SFC and are mapped to the SFF. The residence time and deviation of the residence time
metrics may include any combination of minimum, maximum, values over measurement period,
as well as mean, median, percentile. These metrics may be used to evaluate performance of
the SFC and its elements before and during its operation.
Use of the specially marked packets simplifies residence time measurement
and correlation of the measured metrics over the SFC end-to-end. For example,
the alternate marking method may be used as described in to identify
packets in the data flow to be used to measure the residence time. The nodal and sub-nodal residence time metrics
can be locally calculated and then collected using either in-band or out-band OAM mechanisms.
This document requests IANA to allocate Mark field as two bits-long field from NSH Base Header Reserved Bits .
This document requests IANA to register values of the Mark field of NSH as the following:Bit PositionMarkingDescriptionReference0 S Single Mark Measurement This document1 D Double Mark Measurement This document
This document lists the OAM requirement for SFC domain
and does not raise any security concerns or issues in addition to
ones common to networking and SFC.
TBD