]> Futurewei

2220 Central Expressway Santa Clara CA 95050 USA ludwig@clemm.org

Orange

Rennes 35000 France mohamed.boucadair@orange.com

Ericsson

USA gregimirsky@gmail.com

IPPM This document defines a set of IP Flow Information Export (IPFIX) Information Elements to export precision availability data associated with Flows, specifically Flows that are associated with stringent Service Level Objectives (SLOs) such as latency or packet delay variation.

Introduction IP Flow Information Export (IPFIX) is a protocol that is widely deployed in operators networks to collect Records containing a wide array of statistics about Flows. The Records are used for many purposes, including network security (e.g., detection of denial-of-service attacks), accounting (e.g., identifying "top talkers"), monitoring and service assurance (e.g., detection of anomalies and abnormal behaviors), and network planning (e.g., maintaining traffic matrices and detecting usage trends). To that aim, IPFIX relies upon a set of basic data items that can be maintained by network devices and exported as part of a Flow Record. These data items are commonly referred to as Information Elements (IEs) . Increasingly, to be provided with mere connectivity is no longer sufficient for many networking applications. There is a growing demand for high-precision services that underly stringent Service Level Objectives (SLOs), such as a given latency that must be met by the (connectivity) service. When a guaranteed property of a service (typically, traffic performance metrics) is not met, this is considered in many cases as equivalent to the service not being available. This is particularly the case in which an application relying upon the service does not degrade gracefully with deteriorating service levels (e.g., video or voice), but in which violation of an SLO will cause the application to abruptly cease to function (e.g., industrial control and Control-as-a-Service applications or telehaptics). Existing IPFIX IEs largely focus on statistics such as traffic volume, packet lengths, header fields, or route properties. However, there is a lack of IEs that indicate a Flow's "quality". Specifically, IPFIX does not support IEs that indicate compliance of a Flow with an SLO. This specification fills that void by defining a set of IEs that are based upon Precision Availability Metrics (PAM) . PAMs can thus be exported as part of Flow Records using IPFIX.

Terminology 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. This document uses the IPFIX-specific terminology (Information Element, Template, Collector, Data Record, Flow Record, Exporting Process, Collecting Process, etc.) defined in . As in , these IPFIX-specific terms have the first letter of a word capitalized. Also, this document uses terminology associated with Precision Availability Metrics (PAM), as defined in . For the reader's convenience, some of the acronyms that are used in the document are provided below:

IE:

Information Element

IPFIX:

IP Flow Information Export

PAM:

Precision Availability Metric

SLO:

Service Level Objective

VI:

Violated Interval

VFI:

Violation-Free Interval

Precision Availability Information Elements The following subsections define a set of IEs to export precision availability data as part of Flow Records. At the core of PAMs is the notion of an "interval", i.e. an observation interval (a small unit of time) for which the presence or absence of violations is noted. What constitutes a violation or not depends on the definition of the service, i.e., the length of the interval (e.g., a millisecond) and the SLO (e.g., a not-to-exceed latency threshold or packet inter-arrival delay threshold). Accordingly, IEs are grouped into two categories. The first category contains IEs that reflect PAMs per . The second category contains IEs that are used to define the context that is necessary to adequately interpret the IEs in the first category, such as the SLO that underlies the definition of precision availability for that particular Flow. This context can be thought of as a manifest for that Flow Record.

IEs Based on Precision Availability Metrics

Violated Intervals Count

Name:

violatedIntervalsCount

ElementID:

TBD1

Description:

Contains a count of intervals over the duration of the Flow during which the service was not available with the required precision. That is, a count of intervals for which an SLO violation was observed for the Flow.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Violation-Free Intervals Count

Name:

violationFreeIntervalsCount

ElementID:

TBD2

Description:

Contains a count of intervals over the duration of the Flow during which the required precision was available, i.e., the period during which the Flow was in compliance with its SLO. In practical terms, the violationFreeIntervalsCount corresponds to the number of intervals over the duration of the Flow minus the violatedIntervalsCount.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

• TBD: Assess size of this parameter (for the case of long Flow durations with short interval durations).

Violated Packet Count

Name:

violatedPacketCount

ElementID:

TBD3

Description:

Contains a count of packets for which packet-level violations of an SLO were observed for the Flow.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Severely Violated Intervals Count

Name:

severelyViolatedIntervalsCount

ElementID:

TBD4

Description:

Contains a count of intervals over the duration of a Flow during which a particularly severe violation was observed.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Severely Violated Packet Count

Name:

severelyViolatedPacketCount

ElementID:

TBD5

Description:

Contains a count of packets for which particularly severe packet-level violations of an SLO were observed for the Flow.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Mean Time Between VIs

Name:

meanTimeBetweenViolatedIntervals

ElementID:

TBD6

Description:

Contains the Mean Time Between Violated Intervals over the duration of the Flow.

The mean time is indicated by the number of intervals and thus corresponds to mean number of intervals between violated intervals.

If severelyViolatedIntervalsCount is equal to 0, then the meanTimeBetweenViolatedIntervals must be 0.

If severelyViolatedIntervalsCount is equal to 0, then the meanTimeBetweenViolatedIntervals must be violationFreeIntervalsCount DIV 2.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Mean Number of Packets Between VIs

Name:

meanNumberPacketsBetweenViolatedIntervals

ElementID:

TBD7

Description:

Contains the mean number of packets between packet-level violations over the duration of the Flow.

if violatedPacketCount is equal to 0, then the meanNumberPacketsBetweenViolatedIntervals does not apply.

Abstract Data Type:

unsigned

Data Type Semantics:

quantity

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

• TBD: Which special value to use to indicate that the meanNumberPacketsBetweenViolatedIntervals does not apply.

IEs Representing SLO Manifest Information The following IEs provide context regarding what "violations" and "severe violations" mean for a particular Flow. In this version, IEs for the interval length and for a reference to an SLO are defined. Whether SLOs themselves are to be encoded, including the service level parameter subjected to the SLO (e.g., latency or packet delay variation), the objective itself (upper not-to-exceed threshold or lower threshold and threshold value) is for further study. Likewise, IEs to represent manifest information regarding severity semantics (for severe violations) are for further study.

Precision Availability Interval Length

Name:

precisionAvailabilityIntervalLength

ElementID:

TBD8

Description:

Indicates the duration of an availability interval.

Abstract Data Type:

unsigned

Data Type Semantics:

identifier

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

SLO Identifier

Name:

sloId

ElementID:

TBD9

Description:

A reference to an SLO defining the semantics of what is considered precision availability for the Flow.

Abstract Data Type:

unsigned

Data Type Semantics:

identifier

Additional Information:

See for the general definition of PAM.

Reference:

This-Document

Precision Availability Metrics Not Considered lists a number of additional metrics for which no corresponding IEs are defined for the following reasons:

Time since the last violated interval:

This is a metric that is of interest while a Flow is in progress, but arguably not applicable for export in a Flow Record once the Flow has concluded.

Number of packets since the last violated packet:

By the same token, this is a metric that is of interest while a Flow is in progress, not for export in a Flow Record once the Flow has concluded.

Time since the last severely violated interval:

Analogous reason as for "time since the last violated interval".

Number of packets since the last severely violated packet: :Analogous reason as for "number of packets since the last violated interval".

Mean time between SVIs:

For further study.

Mean packets between SVIs:

For further study.

Violated Interval Ratio:

This can be easily computed by the processor of the Record and does not warrant a separate IE.

Severely Violated Interval Ratio:

This can be easily computed by the processor of the Record and does not warrant a separate IE.

Security Considerations IPFIX security considerations are discussed in .

IANA Considerations This document requests IANA to add the following new IPFIX IEs to the IANA IPFIX registry : New IPFIX Information Elements

Value	Name	Reference
TBD1	violatedIntervalsCount	of This-Document
TBD2	violationFreeIntervalsCount	of This-Document
TBD3	violatedPacketCount	of This-Document
TBD4	severelyViolatedIntervalsCount	of This-Document
TBD5	severelyViolatedPacketCount	of This-Document
TBD6	meanTimeBetweenViolatedIntervals	of This-Document
TBD7	meanNumberPacketsBetweenViolatedIntervals	of This-Document
TBD8	precisionAvailabilityIntervalLength	of This-Document
TBD9	sloId	of This-Document

References Normative References This document specifies the IP Flow Information Export (IPFIX) protocol, which serves as a means for transmitting Traffic Flow information over the network. In order to transmit Traffic Flow information from an Exporting Process to a Collecting Process, a common representation of flow data and a standard means of communicating them are required. This document describes how the IPFIX Data and Template Records are carried over a number of transport protocols from an IPFIX Exporting Process to an IPFIX Collecting Process. This document obsoletes RFC 5101. This document defines the data types and management policy for the information model for the IP Flow Information Export (IPFIX) protocol. This information model is maintained as the IANA "IPFIX Information Elements" registry, the initial contents of which were defined by RFC 5102. This information model is used by the IPFIX protocol for encoding measured traffic information and information related to the traffic Observation Point, the traffic Metering Process, and the Exporting Process. Although this model was developed for the IPFIX protocol, it is defined in an open way that allows it to be easily used in other protocols, interfaces, and applications. This document obsoletes RFC 5102. Ericsson Ericsson ZTE Corp. Futurewei Futurewei Inria and University of Luxembourg This document defines a set of metrics for networking services with performance requirements expressed as Service Level Objectives (SLO). These metrics, referred to as Precision Availability Metrics (PAM), are useful for defining and monitoring SLOs. For example, PAM can be used by providers and/or customers of an RFC XXXX Network Slice Service to assess whether the service is provided in compliance with its defined SLOs. Note to the RFC Editor: Please update "RFC XXXX Network Slice" with the RFC number assigned to draft-ietf-teas-ietf-network-slices. In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. Informative References IANA