Internet Draft Tanja Zseby Document: <draft-ietf-ipfix-as-01.txt> Fraunhofer FOKUS Expires: April 2004 Reinaldo Penno Nortel Networks Nevil Brownlee CAIDA Benoit Claise Cisco Systems October 2003 IPFIX Applicability draft-ietf-ipfix-as-01.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. 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." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. Abstract This document describes what type of applications can use the IP Flow Information Export (IPFIX) protocol and how they can use the information provided by IPFIX. It furthermore shows how the IPFIX framework relates to other architectures and frameworks. Expires April 2004 [Page 1]
IPFIX Applicability October 2003 Table of Contents 1. Introduction.................................................2 2. Applications of IPFIX........................................2 2.1 Accounting...................................................2 2.2 Security Analysis and Intrusion detection with IPFIX.........3 2.3 Network Planning.............................................4 2.4 Peering Agreements...........................................4 2.5 Traffic Engineering..........................................5 2.6 Data Warehousing and Mining..................................5 2.7 Traffic Monitoring...........................................5 2.8 QoS Monitoring...............................................5 2.8.1 Measurement of Round-trip-time (RTT)........................6 2.8.2 Measurement of One-way-delay (OWD)..........................6 2.8.3 Measurement of One-way-loss (OWL)...........................7 2.8.4 Measurement of IP delay variation (IPDV)....................7 3. Relation of IPFIX to other frameworks and protocols..........7 3.1 IPFIX and AAA................................................7 3.1.1 Connecting via an AAA Client................................8 3.1.2 Connecting via an Application Specific Module (ASM).........9 3.2 IPFIX and RTFM..............................................10 3.2.1 Definition of 'flow'.......................................10 3.2.2 Configuration and Management...............................11 3.2.3 Data Model details.........................................12 3.2.4 Application/transport protocol.............................13 3.3 IPFIX and IPPM..............................................13 4. Security Consideration......................................13 5. References..................................................14 6. Acknowledgements............................................15 7. Author's Addresses..........................................15 8. Full Copyright Statement....................................16 1. Introduction The IPFIX protocol defines how IP Flow information can be exported from routers, measurement probes or other devices. It is intended to provide this information as input for various applications. This document describes what applications can use the IPFIX protocol and how they can use it. Furthermore, the relationship of IPFIX to other frameworks and architectures is described. 2. Applications of IPFIX IPFIX data enables several critical customer applications. This section describes how different applications can use IPFIX. 2.1 Accounting Expires April 2004 [Page 2]
IPFIX Applicability October 2003 Usage based accounting is one of the major applications for which the IPFIX protocol has been developed. IPFIX data provides fine-grained metering (for example, flow records include details such as IP addresses, packet and byte counts, timestamps, Type of Service (ToS), application ports, etc.) for highly flexible and detailed resource usage accounting. ISPs can use this information to migrate from single fee, flat-rate billing to more flexible charging mechanisms based on time of day, bandwidth usage, application usage, quality of service, etc. Enterprise customers can use this information for departmental chargeback or cost allocation for resource usage. In order to realize usage-based accounting with IPFIX the flow definition has to be chosen in accordance to the tariff model. A tariff can for instance be based on individual end-to-end streams. In that case accounting can be realized with a flow definition determined by the quintuple that consists of source address, destination address, protocol and portnumbers. Another example is a class-dependent tariff (e.g. in a DiffServ networks). For this flows could be distinguished just by DiffServ codepoint (DSCP) and source address. The essential elements needed for accounting are the number of transferred packets and bytes per flow which are contained in IPFIX flow records. Furthermore IPFIX provides a very flexible definition of flows, so arbitrary flow-based accounting models can be realized without any extensions to the IPFIX protocol. Nevertheless the configuration of flow definitions is out of scope of the IPFIX definition. For accounting purposes, it would be advantageous to have the ability to use IPFIX flow records as accounting input in a AAA infrastructure. AAA servers then could provide the mapping between user and flow information. 2.2 Security Analysis and Intrusion detection with IPFIX Intrusion detection systems (IDS) monitor and control security incidents. A typical IDS system includes components like sensor, event collector, and management stations. Sensors monitor network and system traffic for attacks and other security- related events. Sensors respond to and notify the administrator about these events as they occur. Event collectors are a middle- tier component responsible for transmitting events from sensors to the console and database. The management component serves the following purposes: _ - visually monitors events (with a console) Expires April 2004 [Page 3]
IPFIX Applicability October 2003 _ - collects data from sensors (with one or more event collectors) _ - stores data from sensors (in a database) IPFIX can report events of interest to the sensor either by the collecting process or directly by the exporting process. It depends on the scenario and the events of interest which solution is better. Getting information directly from the exporting process has the advantage that the sensor gets the information faster. It does not need to wait for collector processing time or until the collector has all relevant data. Getting the information from a collector allows correlating data from different exporting processes (e.g. from different routers) to get a better picture about what is going on in the network. IPFIX provides useful input data for basic intrusion detection functions (e.g. detecting unusual high loads) such as details on source and destination addresses, along with the start time of flows, TCP flags, application ports and flow volume. This data can be used to analyze network security and identify attacks. Nevertheless, for some scenarios intrusion detection may require further insight into packet content. Since IPFIX allows a flexible report definition the metering process and the IPFIX report format could be extended to support other data needed for intrusion detection systems. Detecting security incidents in real-time would require a preprocessing of data already at the measurement device and immediate data export in case a possible incident has been identified. This means that IPFIX reports must be generated upon incident detection events and not only upon flow end or fixed time intervals. 2.3 Network Planning IPFIX data captured over a long period of time can be used to track and anticipate network growth and plan upgrades to increase the number of routing devices, ports, or higher- bandwidth interfaces. IPFIX data optimizes both strategic network planning (peering, backbone upgrade planning, and routing policy planning) as well as tactical network engineering decisions (upgrading the router or link capacity). This helps to minimize the total cost of network operations while maximizing network performance, capacity, and reliability. 2.4 Peering Agreements IPFIX data enables ISP peering partners to measure the volume and characteristics of traffic exchanged with other ISP peers. Expires April 2004 [Page 4]
IPFIX Applicability October 2003 2.5 Traffic Engineering IPIFX data provides traffic engineering details for a set of prefixes. This data can be used in network optimization for load balancing traffic across alternate paths, or for forwarding traffic of a certain set of prefixes on a preferred route. 2.6 Data Warehousing and Mining IPFIX data (or derived information) can be stored for later retrieval and analysis to support proactive marketing and customer service programs. An example of this would be to determine which applications and services are being used by internal and external users and then target them for improved services such as advertising. This is especially useful for ISPs because IPFIX data enables them to create better service packaging. 2.7 Traffic Monitoring IPFIX data can be used for extensive near real-time traffic monitoring. Traffic patterns associated with routing devices and switches on an individual or network wide basis can be displayed enabling proactive problem detection, efficient troubleshooting, and rapid problem resolution. IPFIX data enables content and service providers to perform a detailed, time-based, and application-based usage analysis of a network. They also provide detailed information for understanding customer or end-user usage of network and application resources. This information can then be used to efficiently plan and allocate access, backbone, and application resources, as well as to detect and resolve potential security and policy violations. 2.8 QoS Monitoring The performance of QoS monitoring is one target application for using the IPFIX protocol. QoS monitoring is the passive observation of transmission quality for single flows or traffic aggregates in the network. One example of its usefulness is the validation of QoS guarantees in service level agreements (SLAs). Some QoS metrics require the correlation of data from multiple measurement points. For this the clocks of the involved exporting devices must be synchronized. Furthermore, such measurements would benefit from post-processing functions (e.g. packet ID generation and mapping) at the exporter and/or collector. This section describes how the monitoring of Expires April 2004 [Page 5]
IPFIX Applicability October 2003 different metrics can be performed with IPFIX. All of the metrics require at least an extension of the IPFIX information model because currently the necessary information such as e.g. round-trip-time, packet IDs etc. is not part of the model. However given the extensibility and flexibility of IPFIX the missing attributes can be easily defined. 2.8.1 Measurement of Round-trip-time (RTT) The passive measurement of round-trip-times (RTT) can be performed by using packet pair matching techniques as described in [Brow00]. For the measurements, request/response packet pairs from protocols like DNS, ICMP, SNMP or TCP (syn/syn-ack, data/ack) are utilized to passively observe the RTT [Brow00]. As always for passive measurements this only works if the required traffic of interest is actually present in the network. Furthermore, if the observed protocol supports retransmissions (e.g. TCP) the RTT is not the network RTT but rather the RTT of the network and the network stack of the receiver. In case the reply packet is lost or can not be observed the RTT can not be calculated. In order to use this measurement technique, the IPFIX metering process needs to measure both directions. A classification of the protocols mentioned above has to be done. That means parts of the transport header are used for the classification. Since a differentiation of flows in accordance to the transport header is one of the requirements for IPFIX, such classification can be performed without extensions. Nevertheless, the meter needs to recognize request and response packets for the given protocols and therefore needs to look further into the packets. The capability to do this analysis is not part of the IPFIX requirements but can be achieved by optional extensions to the classification process. The exporting device needs to assign a timestamp for the arrival of the packets. The calculation of the RTT can be done directly at the exporter or at the collector. In the first case IPFIX would transfer the calculated RTT to the collector. In the second case IPFIX needs to send the observed packet types and the timestamps to the collector. The round- trip-time-delay metric is defined in [RFC2681]. 2.8.2 Measurement of One-way-delay (OWD) Passive one-way-delay measurements require the collection of data at two measurement points. It is necessary to recognize packets at the second measurement point to correlate packet arrival events from both points. This can be done by capturing packet header and parts of the packet that can be used to recognize the same packet at the subsequent measurement point Expires April 2004 [Page 6]
IPFIX Applicability October 2003 [MaPZ03]. To reduce the amount of measurement data a unique packet ID can be calculated from the header and part and/of the content e.g. by using a CRC or hash function [GrDM98, DuGr00, ZsZC01]. The capability of using content information is out of scope of IPFIX but can be achieved by an optional extension. Nevertheless, in some scenarios it might even be sufficient to calculate a packet ID based on header fields (including datagram ID and maybe sequence numbers from transport protocols) only without looking at parts of the packet content. If packet IDs need to be unique only for a certain time interval or a certain amount of packet ID collisions is tolerable this is a sufficient solution. The second issue is the export of packet IDs. IPFIX exports per flow information. However, it is possible to extend IPFIX with a scheme to export per-packet information by providing special templates for that purpose. The one way delay metric is defined in [RFC2679]. 2.8.3 Measurement of One-way-loss (OWL) Passive loss measurements for single flows can be performed at one measurement point by using sequence numbers that are present in protocols (e.g. IP identification, TCP sequence numbers) similar to the approach described in section 2.8.1. This requires the capturing of the sequence numbers of subsequent packets of the observed flow by the IPFIX metering process. An alternative to this is to perform a two-point measurement as described in section 2.8.2 and consider packets as lost that do not arrive at the second measurement point in a given time frame. This approach assumes that a packet observed at the first point should also be observed at the second point (known routing). The one-way loss metric is defined in [RFC2680]. 2.8.4Measurement of IP delay variation (IPDV) IP Delay variation is defined as the difference of one-way-delay values for selected packets [RFC3393]. Therefore, this metric can be calculated by performing passive measurement of one-way- delay for subsequent packets (e.g. of a flow) and then calculating the differences. 3. Relation of IPFIX to other frameworks and protocols 3.1 IPFIX and AAA AAA defines a protocol and architecture for authentication, authorization and accounting for service usage. The DIAMETER protocol is used for AAA communication for network access Expires April 2004 [Page 7]
IPFIX Applicability October 2003 services (Mobile IP, NASREQ, and ROAMOPS). The AAA architecture [RFC2903] provides a framework for extending the AAA support also for other services. DIAMETER defines the exchange of messages between AAA entities, e.g. between AAA clients at access devices and AAA servers and among AAA servers. It is used also for the transfer of accounting records. Usage-based accounting requires measurement data from the network. IPFIX defines a protocol to export such data from routers, measurement probes and other devices. The provisioning of accounting with IPFIX can be realized without an AAA infrastructure. The collector can directly forward the measurement information to an accounting application. Nevertheless, if an AAA infrastructure is in place, IPFIX can provide the input for the generation of accounting records and several features of the AAA architecture can be used. Features include the mapping of a user ID to the flow information (by using authentication information), the generation of DIAMETER accounting records and the secure exchange of accounting records between domains with DIAMETER. Two possibilities to connect IPFIX and AAA can be distinguished: 3.1.1Connecting via an AAA Client One possibility to connect IPFIX and AAA is to run an AAA client on the IPFIX collector. This client can generate DIAMETER accounting messages and send them to an AAA server. The mapping of the flow information to a user ID can be done in the AAA server by using data from the authentication process. DIAMETER accounting messages can be sent to the accounting application or to other AAA servers (e.g. in roaming scenarios). Expires April 2004 [Page 8]
IPFIX Applicability October 2003 +---------+ DIAMETER +---------+ | AAA-S |------------->| AAA-S | +---------+ +---------+ ^ | DIAMETER | | +--+--------+--+ | | AAA-C | | + +--------+ | | | | Collector | +--------------+ ^ | IPFIX | +------------+ | Exporter | +------------+ Figure 2: IPFIX collector connects to AAA server via AAA client 3.1.2Connecting via an Application Specific Module (ASM) Another possibility is to directly connect the IPFIX collector with the AAA server via an application specific module (ASM). Application specific modules have been proposed by the IRTF AAA architecture research group (AAARCH) in [RFC2903]. They act as an interface between AAA server and service equipment. In this case the IPFIX collector is part of the ASM. The ASM acts as an interface between the IPFIX protocol and the input interface of the AAA server. The ASM translates the received IPFIX data into an appropriate format for the AAA server. The AAA server then can add information about the user ID and generate a DIAMETER accounting record. This accounting record can be sent to an accounting application or to other AAA servers. Expires April 2004 [Page 9]
IPFIX Applicability October 2003 +---------+ DIAMETER +---------+ | AAA-S |------------->| AAA-S | +---------+ +---------+ ^ | +------------------+ | ASM | | +------------+ | | | Collector | | +------------------+ ^ | IPFIX | +------------+ | Exporter | +------------+ Figure 3: IPFIX connects to AAA server via ASM 3.2 IPFIX and RTFM This section compares the Real-time Traffic Flow Measurement (RTFM) framework with the IPFIX framework. 3.2.1Definition of 'flow' RTFM and IPFIX both use the same definition of flow; a flow is a set of packets which share a common set of end-point address attribute values. A flow is therefore completely specified by that set of values, together with an inactivity timeout. A flow is considered to have ended when no packets are seen for at least the inactivity time. RTFM flows are bidirectional, which has given rise to some confusion. At the simplest level, a flow information exporter may achieve this by maintaining two unidirectional flows, one for each direction. To export bidirectional flow information, e.g. to- and from- packet counts, for a flow from A to B, the exporter has only to search its flow table to find the matching flow from B to A. RTFM, however, takes bi-directionality a stage further, by including in the RTFM architecture [RFC 2722] a fully-detailed algorithm for real-time matching of the two directions of a flow. This was done for two reasons, to reduce the memory required to store each flow (common address attributes for each direction), and to allow for attributes which required fine Expires April 2004 [Page 10]
IPFIX Applicability October 2003 detail for the two directions, e.g. short-term bit rate distributions [RFC 2724]. ** So far there has been no suggestion that IPFIX should do this. 3.2.2Configuration and Management The RTFM architecture specifies a complete system for gathering flow information. It defines three entities, - Meters are very similar to IPFIX exporters. - Meter Readers are very similar to IPFIX collectors. - Managers co-ordinate the activities of meters and meter readers, and download configuration to them. Note that the whole RTFM system is asynchronous, many readers may collector flow data from a meter, and any reader may collect flow data from many meters. Rulesets allow the user to specify which flows are of interest, which are the source and destination ends of each flow, and what level of address granularity is required in the metered flows. For example, one may select all packets from 192.168/16, but build flow information for 192.168/24. RTFM selection is done by testing under masks, and the masks do not have to use consecutive ones from the left. Non-contiguous masks were considered important for handling some OSI protocols, but the need for that has diminished considerably. The RTFM approach is based on RMON, in that if a user wants to collect flow data for some particular set of flows, this can be achieved by writing a ruleset, i.e. an SRL program [RFC 2723], to specify what flows are of interest, requesting a manager to download that ruleset to a meter, and requesting the manager to have a meter reader collect the flow data at specified intervals. The details of how the manager communicates this information to meters and meter readers are not specified in the architecture. RTFM has a Meter MIB [RFC 2720], which is a standard which can be used to configure a meter, but nothing is said about how to configure a meter reader. The extent to which IPFIX should specify how meters or exporters should be configured is, at this stage, an open question. Clearly a collector needs some way to be sure of what it's collecting, e.g. by receiving 'templates' from the meter. RTFM and IPFIX both leave parts of the system unspecified. For RTFM flow data to be useful one must know the ruleset used to Expires April 2004 [Page 11]
IPFIX Applicability October 2003 configure the meter, but a user can specify the ruleset. For IPFIX one knows what the data is from the templates, but we have yet to determine whether in-band configuration will be supported. 3.2.3Data Model details 3.2.3.1 Count in one bucket Within a ruleset, a packet may only be counted on one bucket, i.e. it may only be included in one flow. This means that the meter does not have to keep track of overlapping flows - if such aggregation is required, it must be done after the raw flow data has been read by a meter reader. From time to time one may wish to collect flow data for different levels of aggregation at the same time. RTFM allows a meter to run several rulesets at the same time, and meter readers must specify which rulesets they are collecting data from. The 'count in one bucket' rule, together with the ability to run multiple rulesets, has proved very simple and effective in practice. 3.2.3.2 Counter wrapping For its packet- and byte-count attributes RTFM uses continuously-incrementing 64-bit counters, which are never reset. This makes asynchronous meter reading easy, any reader simply has to remember its previous reading and compute the difference. The only caveat is that the meter should be read often enough to avoid situations when the counter has cycled more than once between readings. 3.2.3.3 Sampling issues RTFM provides 1 out of N sampling as a configuration option, so that some metering interfaces may only process every Nth packet. The RTFM Architecture [RFC 2722] does not discuss the statistical implications of this, merely saying that users will need to satisfy themselves that sampling makes sense in their environment. RTFM makes no provision for flow sampling. Recently there has been a lot of interest in flow sampling schemes which favour the 'most important' flows, perhaps we need to consider this for IPFIX. Expires April 2004 [Page 12]
IPFIX Applicability October 2003 3.2.4Application/transport protocol RTFM has a standards-track Meter MIB [RFC 2720], which can be used both to configure a meter and to read flow data from it. The MIB provides a way to read lists of attributes with a single Object Identifier (called a 'package'), which dramatically reduces the SNMP overhead for flow data collection. NeTraMet, a widely-used open-source RTFM implementation, uses SNMPv2C for configuration and data collection. SNMP, of course, normally uses UDP as its transport protocol. Since RTFM requires a reliable flow data transport system, an RTFM meter reader must time out and resend unanswered SNMP requests. Apart from being clumsy, this can limit the maximum data transfer rate from meter to meter reader. SNMP over TCP would be a better approach, but that is currently an IRTF project. On the other hand, RTFM does not specify an application protocol in its architecture, leaving this as an implementation issue. For example, a team at IBM Research implemented a RTFM meter and meter reader in a single host, with the reader storing the flow data directly into a large database system. Similarly, many NeTraMet users run the meter and meter reader on the same host system. A need for high flow data rates highlights the need for careful systems design when building a flow data collection system. When data rates are high, and it is not possible to use a high level of aggregation, then it makes sense to have the collectors very close to their exporters. Once the data is safely on a dedicated host machine, large volumes of it can be moved using 'background' techniques such as FTP. The RTFM architecture only specifies a pull model for getting data out of a meter. To implement push mode data transfer would require specification of triggers to indicate when data should be sent for each flow. 3.3 IPFIX and IPPM The IPFIX protocol can be used to carry IPPM network performance metrics or information which can be used to calculate those metrics (see section 2.8). 4. Security Consideration This document describes the usage of IPFIX in various scenarios. The security requirements for the IPFIX target applications are Expires April 2004 [Page 13]
IPFIX Applicability October 2003 addressed in the IPFIX requirements draft. These requirements must be considered for the specification of the IPFIX protocol. The IPFIX extensions proposed in this document do not induce further security hazards. Section 3 of this document describes how IPFIX can be used in combination with other frameworks. New security hazards can arise when two individually secure frameworks are combined. For the combination of AAA with IPFIX an ASM or an IPFIX collector can function as transit point for the messages. It has to be ensured that at this point the applied security mechanisms (e.g. encryption of messages) are maintained. 5. References [Awdu02] Daniel O. Awduche, et. al.," Overview and Principles of Internet Traffic Engineering", (work in progress), Internet Draft, Internet Engineering Task Force, draft- ietf-tewg-principles-02.txt, May 2002 [Brow00] Nevil Brownlee: Packet Matching for NeTraMet Distributions,http://www2.auckland.ac.nz/net//Internet/r tfm/meetings/47-adelaide/pp-dist/ [DuGr00] Nick Duffield, Matthias Grossglauser: "Trajectory Sampling for Direct Traffic Observation", Proceedings of ACM SIGCOMM 2000, Stockholm, Sweden, August 28 - September 1, 2000. [GrDM98] Ian D. GRAHAM, Stephen F. DONNELLY, Stele MARTIN, Jed MARTENS, John G. CLEARY: Nonintrusive and Accurate Measurement of Unidirectional Delay and Delay Variation on the Internet, INET'98, Geneva, Switzerland, 21-24 July, 1998 [MaPZ03] L. Mark, G. Pohl, T. Zseby, K. Sugauchi: Passive One- way Delay Measurements, (work in progress), Internet Draft <draft-mark-powd-00.txt>, June 2003 [QuZC03] J. Quittek ,et. Al "Requirements for IP Flow Information Export ", (work in progress) ,Internet Draft, Internet Engineering Task Force, <draft-ietf- ipfix-reqs-10.txt>, June 2003 [RFC2679] G. Almes, S. Kalidindi, M. Zekauskas: A One-way Delay Metric for IPPM, Request for Comments: 2679, September 1999 Expires April 2004 [Page 14]
IPFIX Applicability October 2003 [RFC2680] G. Almes, S. Kalidindi, M. Zekauskas: A One-way Packet Loss Metric for IPPM, September 1999 [RFC2681]G. Almes, S. Kalidindi, M. Zekauskas, "A Round-trip Delay Metric for IPPM.", RFC 2681, September 1999 [RFC2903] C. de Laat, G. Gross, L. Gommans, J. Vollbrecht, D. Spence, "Generic AAA Architecture", RFC 2903, August 2000 [RFC3393] C. Demichelis, P. Cimento: IP Packet Delay Variation Metric for IPPM, RFC 3393, November 200 [ZsZC01] Tanja Zseby, Sebastian Zander, Georg Carle: Evaluation of Building Blocks for Passive One-way-delay Measurements, Proceedings of Passive and Active Measurement Workshop (PAM 2001), Amsterdam, The Netherlands, April 23-24, 2001 6. Acknowledgements We would like to thank the following persons for their contribution, discussion on the mailing list and valuable comments: - Sebastian Zander - Robert Loewe 7. Author's Addresses Tanja Zseby Fraunhofer Institute for Open Communication Systems (FOKUS) Kaiserin-Augusta-Allee 31 10589 Berlin, Germany Phone: +49 30 3463 7153 Email: zseby@fokus.fhg.de Reinaldo Penno Nortel Networks, Inc. 2305 Mission College Boulevard Building SC9-B1240, Santa Clara, CA 95134 Email: rpenno@nortelnetworks.com Nevil Brownlee CAIDA (UCSD/SDSC) 9500 Gilman Drive La Jolla, CA 92093-0505 Phone : +1 858 534 8338 Expires April 2004 [Page 15]
IPFIX Applicability October 2003 Email : nevil@caida.org Benoit Claise Cisco Systems De Kleetlaan 6a b1 1831 Diegem Belgium Phone: +32 2 704 5622 Email: bclaise@cisco.com 8. Full Copyright Statement "Copyright (C) The Internet Society (date). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into. Expires April 2004 [Page 16]
