| Internet-Draft | PM with Marking Method in BIER | March 2022 |
| Mirsky, et al. | Expires 2 October 2022 | [Page] |
This document describes the applicability of a hybrid performance measurement method for packet loss and packet delay measurements of a multicast service through a Bit Index Explicit Replication domain.¶
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶
This Internet-Draft will expire on 2 October 2022.¶
Copyright (c) 2022 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
[RFC8279] introduces and explains the Bit Index Explicit Replication (BIER) architecture and how it supports the forwarding of multicast data packets. [RFC8296] specified that in the case of BIER encapsulation in an MPLS network, a BIER-MPLS label, the label that is at the bottom of the label stack, uniquely identifies the multicast flow. [I-D.fioccola-rfc8321bis] and [I-D.fioccola-rfc8889bis] describe a hybrid performance measurement method, according to the classification of measurement methods in [RFC7799]. The method, called Packet Network Performance Monitoring (PNPM), can be used to measure packet loss, latency, and jitter on live traffic complies with requirements R-5 and R-12 listed in [I-D.ietf-bier-oam-requirements]. Because this method is based on marking consecutive batches of packets, the method is often referred to as a marking method. Terms PNPM and "marking method" in this document are used interchangeably.¶
This document defines how the marking method can be used on the BIER layer to measure packet loss and delay metrics of a multicast flow in an MPLS network.¶
This document uses the terms related to the Alternate Marking Method as defined in [I-D.fioccola-rfc8321bis], [I-D.fioccola-rfc8889bis]. This document uses the terms related to the Bit Indexed Explicit Replication as defined in [RFC8296].¶
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 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
[RFC8296] defined the two-bits long field, referred to as OAM. The OAM field can be used for the marking performance measurement method. Because the setting of the field to any value does not affect forwarding and/or quality of service treatment of a packet, using the OAM field for PNPM in BIER layer can be viewed as the example of the hybrid performance measurement method.¶
Figure 1 displays the interpretation of the OAM field defined in this specification for the use of the PNPM method. The context of interpretation of the OAM field MAY be signaled via the control plane or configured using an extension to the BIER YANG data model [I-D.ietf-bier-bier-yang]. These extensions are outside the scope of this document.¶
0 0 1 +-+-+-+-+ | S | D | +-+-+-+-+
where:¶
The marking method can be used in the multicast environment supported by BIER layer. Without limiting any generality consider multicast network presented in Figure 2. Any combination of markings can be applied to a multicast flow by the Bit Forwarding Ingress Router (BFIR) 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.¶
-----
--| D |
----- / -----
--| B |--
/ ----- \ -----
/ --| E |
----- / -----
| A |--- -----
----- \ --| F |
\ ----- / -----
--| C |--
----- \ -----
--| G |
-----
Using the marking method, a BFIR creates distinct sub-flows in the particular multicast traffic over BIER layer. Each sub-flow consists of consecutive blocks of identically marked packets. For example, a block of N packets, with each packet being marked as X, is followed by the block of M packets with each packet being marked as Y. These blocks are unambiguously recognizable by a monitoring point at any Bit Forwarding Router (BFR) and can be measured to calculate packet loss and/or packet delay metrics. The marking method can be used on multiple flows concurently. Demultiplexing of monitored flows might be achived using n-tuple, for example, two-tuple as combination of the values in the Entropy and BFIR-id fields [RFC8296]. Also, that can be achieved by using an explicit Flow Identifiier. The definition of the Flow Identifier is outside the scope of this specification. It is expected that the marking values be set and cleared at the edge of BIER domain. Thus for the scenario presented in Figure 2 if the operator initially monitors the A-C-G and A-B-D segments he may enable measurements on segments C-F and B-E at any time.¶
As explained in [I-D.fioccola-rfc8321bis], marking can be applied to delineate blocks of packets based either on the equal number of packets in a block or based on the equal time interval. The latter method offers better control as it allows a 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-Marking measurement is used to measure packet loss, then the D flag MUST be set to zero on transmit and ignored by the monitoring point.¶
The S flag is used to create sub-flows to measure the packet loss by switching the value of the S flag every N-th packet or at certain time intervals. Delay metrics MAY be calculated with the sub-flow using any of the following methods:¶
Double-Marking method allows measurement of minimum and maximum delays for the monitored flow, but it requires more nodal and network resources. If the Double-Marking method used, then the S flag is used to create the sub-flow, i.e., mark blocks of packets. The D flag is used to mark single packets within a block to measure delay and jitter.¶
The first marking (S flag alternation) is needed for packet loss and also for average delay measurement. The second marking (D flag is put to one) creates a new set of marked packets that are fully identified over the BIER network, so that a BFR can store the timestamps of these packets; these timestamps can be compared with the timestamps of the same packets on a second BFR to compute packet delay values for each packet. The number of measurements can be easily increased by changing the frequency of the second marking. On the other hand, the higher frequency of the second marking will cause a higher volume of the measurement data being transported through the BIER domain. An operator should consider and balance both effects. This method is useful to measure 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.¶
For the ease of operational procedures, the initial marking of a multicast flow is performed at BFIR. and cleared, by way of removing BIER encapsulation form a payload packet, at the edge of the BIER domain by BFERs.¶
Since at the time of writing this specification, there are no proposals to using auto-discovery or signaling mechanism to inform downstream nodes what methodology is used each monitoring point MUST be configured beforehand.¶
Section 5 [I-D.fioccola-rfc8321bis] provides a detailed analysis of how packet re-ordering and the duration of the block in the Single-Marking mode of the marking method impact the accuracy of the packet loss measurement. Re-ordering of packets in the Single-Marking mode will be noticeable only at the edge of a block of packets (re-ordering within the block cannot be detected in the Single-Marking mode). If the extra delay for some packets is much smaller than half of the duration of a block, then it should be easier to attribute re-ordered packets to the proper block and thus maintain the accuracy of the packet loss measurement.¶
Selection of a time interval to switch the marking of a batch of packets should be based on the service requirements. In the course of the regular operation, reports, including performance metrics like packet loss ratio, packet delay, and inter-packet delay variation, are logged every 15 minutes. Thus, it is reasonable to maintain the duration of the measurement interval at 5 minutes with 100 measurements per each interval. To support these measurements, marking of the packet batch is switched every 3 seconds. In case when performance metrics are required in near-real-time, the duration interval of a single batch of identically marked packets will be in the range of tens of milliseconds.¶
This document sets no requirements to IANA. This section can be removed before the publication.¶
Regarding using the marking method, [I-D.fioccola-rfc8321bis] stressed two types of security concerns. First, the potential harm caused by the measurements, is a lesser threat as [RFC8296] defines OAM field used by the marking method so that the value of "two bits have no effect on the path taken by a BIER packet and have no effect on the quality of service applied to a BIER packet." Second security concern, potential harm to the measurements can be mitigated by using policy, suggested in [RFC8296], to accept BIER packets only from trusted routers, not from customer-facing interfaces.¶
All the security considerations for BIER discussed in [RFC8296] are inherited by this document.¶
Comments from Alvaro Retana helped improve the document and are much appreciated.¶
Reviews and comments from Quan Xiong and Xiao Min are thankfully acknowledged.¶