IPv6 Updates for IPPM's Active Metric Framework
draft-ietf-ippm-2330-ipv6-01

Abstract

This memo updates the IP Performance Metrics (IPPM) Framework RFC 2330 with new considerations for measurement methodology and testing. It updates the definition of standard-formed packets in RFC 2330 to include IPv6 packets and augments distinguishing aspects of packets, referred to as Type-P for test packets in RFC 2330. This memo identifies that IPv4-IPv6 co-existence can challenge measurements within the scope of the IPPM Framework. Exemplary use cases include, but are not limited to IPv4-IPv6 translation, NAT, protocol encapsulation, or IPv6 header compression.

Requirements Language

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].

Status of This Memo

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 http://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 September 7, 2017.

Copyright Notice

Copyright (c) 2017 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 (http://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 Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

Table of Contents

• 1. Introduction
• 2. Scope
• 3. Packets of Type-P
• 4. Standard-Formed Packets
• 5. NAT, IPv4-IPv6 Transition and Compression Techniques
• 6. Security Considerations
• 7. IANA Considerations
• 8. Acknowledgements
• 9. References
• 9.1. Normative References
• 9.2. Informative References
• Authors' Addresses

1. Introduction

The IETF IP Performance Metrics (IPPM) working group first created a framework for metric development in [RFC2330]. This framework has stood the test of time and enabled development of many fundamental metrics. It has been updated in the area of metric composition [RFC5835], and in several areas related to active stream measurement of modern networks with reactive properties [RFC7312].

The IPPM framework [RFC2330] recognized (in section 13) that many aspects of IP packets can influence its processing during transfer across the network.

In Section 15 of [RFC2330], the notion of a "standard-formed" packet is defined. However, the definition was never updated to include IPv6, as the original authors planned.

In particular, IPv6 Extension Headers and protocols which use IPv6 header compression are growing in use. This memo seeks to provide the needed updates.

2. Scope

The purpose of this memo is to expand the coverage of IPPM metrics to include IPv6, and to highlight additional aspects of test packets and make them part of the IPPM performance metric framework.

The scope is to update key sections of [RFC2330], adding considerations that will aid the development of new measurement methodologies intended for today's IP networks. Specifically, this memo expands the Type-P examples in section 13 of [RFC2330] and expands the definition (in section 15 of [RFC2330]) of a standard-formed packet to include IPv6 header aspects and other features.

Other topics in [RFC2330] which might be updated or augmented are deferred to future work. This includes the topics of passive and various forms of hybrid active/passive measurements.

3. Packets of Type-P

A fundamental property of many Internet metrics is that the measured value of the metric depends on characteristics of the IP packet(s) used to make the measurement. Potential influencing factors include IP header fields and their values, but also higher-layer protocol headers and their values. Consider an IP-connectivity metric: one obtains different results depending on whether one is interested in connectivity for packets destined for well-known TCP ports or unreserved UDP ports, or those with invalid IPv4 checksums, or those with TTL or Hop Limit of 16, for example. In some circumstances these distinctions will result in special treatment of packets in intermediate nodes and end systems (for example, if Diffserv [RFC2780], ECN [RFC3168], Router Alert, Hop-by-hop extensions [RFC7045], or Flow Labels [RFC6437] are used, or in the presence of firewalls or RSVP reservations).

Because of this distinction, we introduce the generic notion of a "packet of Type-P", where in some contexts P will be explicitly defined (i.e., exactly what type of packet we mean), partially defined (e.g., "with a payload of B octets"), or left generic. Thus we may talk about generic IP-Type-P-connectivity or more specific IP-port-HTTP-connectivity. Some metrics and methodologies may be fruitfully defined using generic Type-P definitions which are then made specific when performing actual measurements.

Whenever a metric's value depends on the type of the packets involved in the metric, the metric's name will include either a specific type or a phrase such as "Type-P". Thus we will not define an "IP-connectivity" metric but instead an "IP-Type-P-connectivity" metric and/or perhaps an "IP-port-HTTP-connectivity" metric. This naming convention serves as an important reminder that one must be conscious of the exact type of traffic being measured.

If the information constituting Type-P at the Source is found to have changed at the Destination (or at a measurement point between the Source and Destination, as in [RFC5644]), then the modified values SHOULD be noted and reported with the results. Some modifications occur according to the conditions encountered in transit (such as congestion notification) or due to the requirements of segments of the Source to Destination path. For example, the packet length will change if IP headers are converted to the alternate version/address family, or if optional Extension Headers are added or removed. Even header fields like TTL/Hop Limit that typically change in transit may be relevant to specific tests. For example Neighbor Discovery Protocol (NDP) [RFC4861] packets are transmitted with Hop Limit value set to 255, and the validity test specifies that the Hop Limit must have a value of 255 at the receiver, too. So, while other tests may intentionally exclude the TTL/Hop Limit value from their Type-P definition, for this particular test the correct Hop Limit value is of high relevance and must be part of the Type-P definition.

Local policies in intermediate nodes based on examination of IPv6 Extension Headers may affect measurement repeatability. If intermediate nodes follow the recommendations of [RFC7045], repeatability may be improved to some degree.

A closely related note: it would be very useful to know if a given Internet component (like host, link, or path) treats equally a class C of different types of packets. If so, then any one of those types of packets can be used for subsequent measurement of the component. This suggests we devise a metric or suite of metrics that attempt to determine C.

Load balancing over parallel paths is one particular example where such a class C would be more complex to determine in IPPM measurements. Load balancers often use flow identifiers, computed as hashes of (specific parts of) the packet header, for deciding among the available parallel paths a packet will traverse. Packets with identical hashes are assigned to the same flow and forwarded to the same resource in the load balancer's pool. The presence of a load balancer on the measurement path, as well as the specific headers and fields that are used for the forwarding decision, are not known when measuring the path as a black-box. Potential assessment scenarios include the measurement of one of the parallel paths, and the measurement of all available parallel paths that the load balancer can use. Knowledge of a load balancer's flow definition (alternatively: its class C specific treatment in terms of header fields in scope of hash operations) is therefore a prerequisite for repeatable measurements. A path may have more than one stage of load balancing, adding to class C definition complexity.

4. Standard-Formed Packets

Unless otherwise stated, all metric definitions that concern IP packets include an implicit assumption that the packet is *standard-formed*. A packet is standard-formed if it meets all of the following criteria:

( [RFC0791] and updates) packet to be standard-formed, the following additional criteria are required:

+

It includes a valid IP header: see below for version-specific criteria.

+

It is not an IP fragment.

+

The Source and Destination addresses correspond to the intended Source and Destination, including Multicast Destination addresses.

+

If a transport header is present, it contains a valid checksum and other valid fields.

For an IPv4

• The version field is 4
• The Internet Header Length (IHL) value is >= 5; the checksum is correct.
• Its total length as given in the IPv4 header corresponds to the size of the IPv4 header plus the size of the payload.
• Either the packet possesses sufficient TTL to travel from the Source to the Destination if the TTL is decremented by one at each hop, or it possesses the maximum TTL of 255.
• It does not contain IP options unless explicitly noted.

For an IPv6 ([RFC2460] and updates) packet to be standard-formed, the following criteria are required:

• The version field is 6.
• Its total length corresponds to the size of the IPv6 header (40 octets) plus the length of the payload as given in the IPv6 header.
• The payload length value for this packet (including Extension Headers) conforms to the IPv6 specifications.
• Either the packet possesses sufficient Hop Count to travel from the Source to the Destination if the Hop Count is decremented by one at each hop, or it possesses the maximum Hop Count of 255.
• Either the packet does not contain IP Extension Headers, or it contains the correct number and type of headers as specified in the packet, and the headers appear in the standard-conforming order (Next Header).
• All parameters used in the header and Extension Headers are found in the IANA Registry of Internet Protocol Version 6 (IPv6) Parameters, partly specified in [RFC7045].

Compressed IPv6 headers must be compliant with [RFC4494], as updated by [RFC6282], in order to be declared "standard-formed".

The topic of IPv6 Extension Headers brings current controversies into focus as noted by [RFC6564] and [RFC7045]. The following additional considerations apply when IPv6 Extension Headers are present:

[RFC2675] are not covered by this length analysis. In practice, the path MTU will restrict the length of standard-formed packets that can successfully traverse the path.

• Extension Header inspection: Some intermediate nodes may inspect Extension Headers or the entire IPv6 packet while in transit. In exceptional cases, they may drop the packet or route via a sub-optimal path, and measurements may be unreliable or unrepeatable. The packet (if it arrives) may be standard-formed, with a corresponding Type-P.
• Extension Header modification: In Hop-by-Hop headers, some TLV encoded options may be permitted to change at intermediate nodes while in transit. The resulting packet may be standard-formed, with a corresponding Type-P.
• Extension Header insertion or deletion: It is possible that Extension Headers could be added to, or removed from the header chain. The resulting packet may be standard-formed, with a corresponding Type-P.
• A change in packet length (from the corresponding packet observed at the Source) or header modification is a significant factor in Internet measurement, and requires a new Type-P to be reported.

We further require that if a packet is described as having a "length of B octets", then 0 <= B <= 65535; and if B is the payload length in octets, then B <= (65535-IP header size in octets, including any Extension Headers). The jumbograms defined in

So, for example, one might imagine defining an IP connectivity metric as "IP-type-P-connectivity for standard-formed packets with the IP Diffserv field set to 0", or, more succinctly, "IP-type-P-connectivity with the IP Diffserv Field set to 0", since standard-formed is already implied by convention. Changing the contents of a field, such as the Diffserv Code Point, ECN bits, or Flow Label may have a profound affect on packet handling during transit, but does not affect a packet's status as standard-formed.

A particular type of standard-formed packet often useful to consider is the "minimal IP packet from A to B" - this is an IP packet with the following properties:

+

It is standard-formed.

+

Its data payload is 0 octets.

+

It contains no options or Extension Headers.

(Note that we do not define its protocol field, as different values may lead to different treatment by the network.)

When defining IP metrics we keep in mind that no packet smaller or simpler than this can be transmitted over a correctly operating IP network.

5. NAT, IPv4-IPv6 Transition and Compression Techniques

This memo adds the key considerations for utilizing IPv6 in two critical conventions of the IPPM Framework, namely packets of Type-P and standard-formed packets. The need for co-existence of IPv4 and IPv6 has originated transitioning standards like the Framework for IPv4/IPv6 Translation in [RFC6144] or IP/ICMP Translation Algorithms in [RFC6145] and [RFC7757].

The definition and execution of measurements within the context of the IPPM Framework is challenged whenever such translation mechanisms are present along the measurement path. In particular use cases like IPv4-IPv6 translation, NAT, protocol encapsulation, or IPv6 header compression may result in modification of the measurement packet's Type-P along the path. All these changes must be reported. Exemplary consequences include, but are not limited to:

• Modification or addition of headers or header field values in intermediate nodes. As noted in Section 4 for IPv6 extension header manipulation, NAT, IPv4-IPv6 transitioning or IPv6 header compression mechanisms may result in changes of the measurement packets' Type-P, too. Consequently, hosts along the measurement path may treat packets differently because of the Type-P modification. Measurements at observation points along the path may also need extra context to uniquely identify a packet.
• Network Address Translators (NAT) on the path can have unpredictable impact on latency measurement (in terms of the amount of additional time added), and possibly other types of measurements. It is not usually possible to control this impact (as testers may not have any control of the underlying network or middleboxes). There is a possibility that stateful NAT will lead to unstable performance for a flow with specific Type-P, since state needs to be created for the first packet of a flow, and state may be lost later if the NAT runs out of resources. However, this scenario does not invalidate the Type-P for testing - for example the purpose of a test might be exactly to quantify the NAT's impact on delay variation. The presence of NAT may mean that the measured performance of Type-P will change between the source and the destination. This can cause an issue when attempting to correlate measurements conducted on segments of the path that include or exclude the NAT. Thus, it is a factor to be aware of when conducting measurements.
• Variable delay due to internal state. One side effect of changes due to IPv4-IPv6 transitioning mechanisms is the variable delay that intermediate nodes spend for header modifications. Similar to NAT the allocation of internal state and establishment of context within intermediate nodes may cause variable delays, depending on the measurement stream pattern and position of a packet within the stream. For example the first packet in a stream will typically trigger allocation of internal state in an intermediate IPv4-IPv6 transition host. Subsequent packets can benefit from lower processing delay due to the existing internal state. However, large inter-packet delays in the measurement stream may result in the intermediate host deleting the associated state and needing to re-establish it on arrival of another stream packet. It is worth noting that this variable delay due to internal state allocation in intermediate nodes can be an explicit use case for measurements.
• Variable delay due to packet length. IPv4-IPv6 transitioning or header compression mechanisms modify the length of measurement packets. The modification of the packet size may or may not change the way how the measurement path treats the packets.

Points that are worthwhile discussing further: handling of large packets in IPv6 (including fragment extension headers, PMTUD, PLMTUD), extent of coverage for 6LO and IPv6 Header Compression, and the continued need to define a "minimal standard-formed packet".

.

6. Security Considerations

The security considerations that apply to any active measurement of live paths are relevant here as well. See [RFC4656] and [RFC5357].

When considering privacy of those involved in measurement or those whose traffic is measured, the sensitive information available to potential observers is greatly reduced when using active techniques which are within this scope of work. Passive observations of user traffic for measurement purposes raise many privacy issues. We refer the reader to the privacy considerations described in the Large Scale Measurement of Broadband Performance (LMAP) Framework [RFC7594], which covers active and passive techniques.

7. IANA Considerations

This memo makes no requests of IANA.

8. Acknowledgements

The authors thank Brian Carpenter for identifying the lack of IPv6 coverage in IPPM's Framework, and for listing additional distinguishing factors for packets of Type-P. Both Brian and Fred Baker discussed many of the interesting aspects of IPv6 with the co-authors, leading to a more solid first draft: thank you both. Thanks to Bill Jouris for an editorial pass through the pre-00 text.

9. References

9.1. Normative References

9.2. Informative References

Authors' Addresses