idnits 2.17.1 draft-ietf-rtfm-new-traffic-flow-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Looks like you're using RFC 2026 boilerplate. This must be updated to follow RFC 3978/3979, as updated by RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** Missing expiration date. The document expiration date should appear on the first and last page. == No 'Intended status' indicated for this document; assuming Proposed Standard Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an Abstract section. ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) == There are 1 instance of lines with non-RFC6890-compliant IPv4 addresses in the document. If these are example addresses, they should be changed. Miscellaneous warnings: ---------------------------------------------------------------------------- == Line 260 has weird spacing: '...r every packe...' == Line 445 has weird spacing: '...e could be pu...' -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- Couldn't find a document date in the document -- date freshness check skipped. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Missing reference section? '1' on line 670 looks like a reference -- Missing reference section? '4' on line 680 looks like a reference -- Missing reference section? '2' on line 674 looks like a reference -- Missing reference section? '8' on line 694 looks like a reference -- Missing reference section? '3' on line 677 looks like a reference -- Missing reference section? '6' on line 688 looks like a reference -- Missing reference section? '9' on line 697 looks like a reference -- Missing reference section? '7' on line 691 looks like a reference -- Missing reference section? '5' on line 684 looks like a reference Summary: 4 errors (**), 0 flaws (~~), 4 warnings (==), 11 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 1 Real Time Flow Measurement Working Group S.W. Handelman 2 Internet-draft IBM 3 Hawthorne, NY USA 5 Nevil Brownlee 6 U of Auckland, NZ 8 Greg Ruth 9 GTE Laboratories, Inc 10 Waltham, MA USA 12 S. Stibler 13 IBM 14 Hawthorne, NY USA 16 April, 1999 17 expires 18 October, 1999 20 RTFM Working Group - New Attributes for Traffic Flow Measurement 22 24 1. Status of this Memo 26 This document is an Internet-Draft and is in full conformance with 27 all provisions of Section 10 of RFC2026. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF), its areas, and its working groups. Note that 31 other groups may also distribute working documents as Internet- 32 Drafts. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 The list of current Internet-Drafts can be accessed at 40 http://www.ietf.org/ietf/1id-abstracts.txt 42 The list of Internet-Draft Shadow Directories can be accessed at 43 http://www.ietf.org/shadow.html. 45 2. Introduction 47 The Real-Time Flow Measurement (RTFM) Working Group (WG) has 48 developed a system for measuring and reporting information about 49 traffic flows in the Internet. This document explores the definition 50 of extensions to the flow measurements as currently defined in [1]. 51 The new attributes described in this document will be useful for 52 monitoring network performance and will expand the scope of RTFM 53 beyond simple measurement of traffic volumes. A companion document 54 to this draft will be written to define MIB structures for the new 55 attributes. 57 This draft was started in 1996 to advance the work of the RTFM group. 58 The goal of this work is to produce a simple set of abstractions, 59 which can be easily implemented and at the same time enhance the 60 value of RTFM Meters. This document also defines a method for 61 organizing the flow abstractions to augment the existing RTFM flow 62 table. 64 Implementations of the RTFM Meter have been done by Nevil Brownlee in 65 the University of Auckland, NZ, and Stephen Stibler and Sig Handelman 66 at IBM in Hawthorne, NY, USA. The RTFM WG has also defined the role 67 of the Meter Reader whose role is to retrieve flow data from the 68 Meter. 70 Note on flows and positioning of meters: 72 A flow as it traverses the Internet may have some of its 73 characteristics altered as it travels through Routers, Switches, and 74 other network units. It is important to note the spatial location of 75 the Meter when referring to attributes of a flow. An example, a 76 server may send a sequence of packets with a definite order, and 77 inter packet timing with a leaky bucket algorithm. A meter reading 78 downstream of the leaky bucket would record a set with minimal inter 79 packet timing due to the leaky bucket. At the client's location, the 80 packets may arrive out of sequence, with the timings altered. A meter 81 at the client's location would record different attributes for the 82 same flow. 84 2.1. RTFM's Definition of Flows 86 The RTFM Meter architecture views a flow as a set of packets between 87 two endpoints (as defined by their source and destination attribute 88 values and start and end times), and as BI-DIRECTIONAL (i.e. the 89 meter effectively monitors two sub-flows, one in each direction). 91 Reasons why RTFM flows are bi-directional: 93 - The WG is interested in understanding the behavior of sessions 94 between endpoints. 96 - The endpoint attribute values (the "Address" and "Type" ones) are 97 the same for both directions; storing them in bi-directional flows 98 reduces the meter's memory demands. 100 - 'One-way' (uni-directional) flows are a degenerate case. 101 Existing RTFM meters can handle this by using one of the computed 102 attributes (e.g. FlowKind) to indicate direction. 104 2.2. RTFM's Current Definition of Flows and their Attributes 106 Flows, as described in the "Architecture" document [1] have the 107 following properties: 109 a. They occur between two endpoints, specified as sets of attribute 110 values in the meter's current rule set. A flow is completely 111 identified by its set of endpoint attribute values. 113 b. Each flow may also have values for "computed" attributes (Class 114 and Kind). These are directly derived from the endpoint 115 attribute values. 117 c. A new flow is created when a packet is to be counted that does 118 not match the attributes of an existing flow. The meter records 119 the time when this new flow is created. 121 d. Attribute values in (a), (b) and (c) are set when the meter sees 122 the first packet for the flow, and are never changed. 124 e. Each flow has a "LastTime" attribute, which indicates the time 125 the meter last saw a packet for the flow. 127 f. Each flow has two packet and two byte counters, one for each 128 flow direction (Forward and Backward). These are updated as 129 packets for the flow are observed by the meter. 131 g. ALL the attributes have (more or less) the same meaning for a 132 variety of protocols; IPX, AppleTalk, DECnet and CLNS as well 133 as TCP/IP. 135 Current flow attributes - as described above - fit very well into the 136 SNMP data model. They are either static, or are continuously updated 137 counters. They are NEVER reset. In this document they will be 138 referred to as "old-style" attributes. 140 It is easy to add further "old-style" attributes, since they don't 141 require any new features in the architecture. For example: 143 - Count of the number of "lost" packets (determined by watching 144 sequence number fields for packets in each direction; only 145 available for protocols which have such sequence numbers). 147 - In the future, RTFM could coordinate directly with the Flow Label 148 from the IPv6 header. 150 2.3. RTFM Flows, Integrated Services, IPPM and Research in Flows 152 The concept of flows has been studied in various different contexts. 153 For the purpose of extending RTFM, a starting point is the work of 154 the Integrated Services WG. We will measure quantities that are often 155 set by Integrated Services configuration programs. We will look at 156 the work of the Benchmarking / IP Performance Metrics Working Group, 157 and also look at the work of Claffy, Braun and Polyzos [4]. We will 158 demonstrate how RTFM can compute throughput, packet loss, and delays 159 from flows. 161 An example of the use of capacity and performance information is 162 found in "The Use of RSVP with IETF Integrated Services" [2]. RSVP's 163 use of Integrated Services revolves around Token Bucket Rate, Token 164 Bucket Size, Peak Data Rate, Minimum Policed Unit, Maximum Packet 165 Size, and the Slack term. These are set by TSpec, ADspec and FLowspec 166 (Integrated Services Keywords), and are used in configuration and 167 operation of Integrated Services. RTFM could monitor explicitly Peak 168 Data Rate, Minimum Policed Unit, Maximum Packet Size, and the Slack 169 term. RTFM could infer details of the Token Bucket. The WG will 170 develop measures to work with these service metrics. An initial 171 implementation of IIS Monitoring has been developd at CEFRIEL in 172 Italy [8]. 174 RTFM will work with several traffic measurements identified by IPPM 175 [3]. There are three broad areas in which RTFM is useful for IPPM. 177 - An RTFM Meter could act as a passive device, gathering traffic 178 and performance statistics at appropriate places in networks 179 (server or client locations). 181 - RTFM could give detailed analyses of IPPM test flows that pass 182 through the Network segment that RTFM is monitoring. 184 - RTFM could be used to identify the most-used paths in a network 185 mesh, so that detailed IPPM work could be applied to these most 186 used paths. 188 3. Flow Abstractions 190 Performance attributes include throughput, packet loss, delays, 191 jitter, and congestion measures. RTFM will calculate these attributes 192 in the form of extensions to the RTFM flow attributes according to 193 three general classes: 195 - 'trace,' attributes of individual packets in a flow or a segment 196 of a flow (e.g. last packet size, last packet arrival time). 198 - 'aggregate,' attributes derived from the flow taken as a whole 199 (e.g. mean rate, max packet size, packet size distribution). 201 - 'group,' attributes that depend on groups of packet values within 202 the flow (e.g. inter-arrival times, short-term traffic rates). 204 Note that attributes within each of these classes may have various 205 types of values - numbers, distributions, time series, and so on. 207 3.1. Meter Readers and Meters 209 A note on the relation between Meter Readers and Meters. 211 Several of the measurements enumerated below can be implemented by a 212 Meter Reader that is tied to a meter with very short response time 213 and very high bandwidth. If the Meter Reader and Meter can be 214 arranged in such a way, RTFM could collect Packet Traces with time 215 stamps and provide them directly to the Meter Reader for further 216 processing. 218 A more useful alternative is to have the Meter calculate some flow 219 statistics locally. This allows a looser coupling between the Meter 220 and Meter Reader. RTFM will monitor an 'extended attribute' depending 221 upon settings in its Rule table. RTFM will not create any "extended 222 attribute" data without explicit instructions in the Rule table. 224 3.2. Attribute Types 226 Section 3. described three different classes of attributes; this 227 section considers the "data types" of these attributes. 229 Packet Traces (as described below) are a special case in that they 230 are tables with each row containing a sequence of values, each of 231 varying type. They are essentially 'compound objects' i.e. lists of 232 attribute values for a string of packets. 234 Aggregate attributes are like the 'old-style' attributes. Their 235 types are 237 - Addresses, represented as byte strings (1 to 20 bytes long) 239 - Counters, represented as 64-bit unsigned integers 241 - Times, represented as 32-bit unsigned integers 243 Addresses are saved when the first packet of a flow is observed. They 244 do not change with time, and they are used as a key to find the 245 flow's entry in the meter's flow table. 247 Counters are incremented for each packet, and are never reset. An 248 analysis application can compute differences between readings of the 249 counters, so as to determine rates for these attributes. For 250 example, if we read flow data at five-minute intervals, we can 251 calculate five-minute packet and byte rates for the flow's two 252 directions. 254 Times - the FirstTime for a flow is set when its first packet is 255 observed. LastTime is updated as each packet in the flow is observed. 257 All the above types have the common feature that they are expressed 258 as single values. At least some of the new attributes will require 259 multiple values. If, for example, we are interested in inter-packet 260 time intervals, we can compute an interval for every packet after 261 the first. If we are interested in packet sizes, a new value is 262 obtained as each packet arrives. When it comes to storing this data 263 we have two options: 265 - As a distribution, i.e. in an array of 'buckets.' This method is 266 a compact representation of the data, with the values being stored 267 as counters between a minimum and maximum, with defined steps in 268 each bucket. This fits the RTFM goal of compact data storage. 270 - As a sequence of single values. This saves all the information, 271 but does not fit well with the RTFM goal of doing as much data 272 reduction as possible within the meter. 274 Studies which would be limited by the use of distributions might well 275 use packet traces instead. 277 A method for specifying the distribution parameters, and for encoding 278 the distribution so that it can be easily read, is described in 279 section 4.2. 281 3.3. Packet Traces 283 The simplest way of collecting a trace in the meter would be to have 284 a new attribute called, say, "PacketTrace." This could be a table, 285 with a column for each property of interest. For example, one could 286 trace 288 - Packet Arrival time (TimeTicks from sysUpTime, or microseconds 289 from FirstTime for the flow). 291 - Packet Direction (Forward or Backward) 293 - Packet Sequence number (for protocols with sequence numbers) 295 - Packet Flags (for TCP at least) 297 Note: The following implementation proposal is for the user who is 298 familiar with the writing of rule sets for the RTFM Meter. 300 To add a row to the table, we only need a rule which PushPkts the 301 PacketTrace attribute. To use this, one would write a rule set which 302 selected out a small number of flows of interest, with a 'PushPkt 303 PacketTrace' rule for each of them. A MaxTraceRows default value of 304 2000 would be enough to allow a Meter Reader to read one-second ping 305 traces every 10 minutes or so. More realistically, a MaxTraceRows of 306 500 would be enough for one-minute pings, read once each hour. Note 307 that packet traces are already implemented in the RMON MIB [6], in 308 the Packet Capture Group. They are therefore a low priority for RTFM. 310 3.4. Aggregate Attributes 312 RTFM's "old-style" flow attributes count the bytes and packets for 313 packets which match the rule set for an individual flow. In addition 314 to these totals, for example, RTFM could calculate Packet Size 315 statistics. This data can be stored as distributions, though it may 316 sometimes be sufficient to simply keep a maximum value. 318 Packet Size - RTFM's packet flows can be examined to determine the 319 maximum packet size found in a flow. This will give the Network 320 Operator an indication of the MTU being used in a flow. It will also 321 give an indication of the sensitivity to loss of a flow, for losing 322 large packets causes more data to be retransmitted. 324 Note that aggregate attributes are a simple extension of the 'old- 325 style' attributes; their values are never reset. For example, an 326 array of counters could hold a 'packet size' distribution. The 327 counters continue to increase, a meter reader will collect their 328 values at regular intervals, and an analysis application will compute 329 and display distributions of the packet size for each collection 330 interval. 332 3.5. Group Attributes 334 The notion of group attributes is to keep simple statistics for 335 measures that involve more than one packet. This section describes 336 some group attributes which it is feasible to implement in a traffic 337 meter, and which seem interesting and useful. 339 Short-term bit rate - The data could also be recorded as the maximum 340 and minimum data rate of the flow, found over specific time periods 341 during the lifetime of a flow; this is a special kind of 342 'distribution.' Bit rate could be used to define the throughput of a 343 flow, and if the RTFM flow is defined to be the sum of all traffic in 344 a network, one can find the throughput of the network. 346 If we are interested in '10-second' forward data rates, the meter 347 might compute this for each flow of interest as follows: 349 - maintain an array of counters to hold the flow's 10-second 350 data rate distribution. 352 - every 10 seconds, compute and save 10-second octet count, and 353 save a copy of the flow's forward octet counter. 355 To achieve this, the meter will have to keep a list of aggregate 356 flows and the intervals at which they require processing. Careful 357 programming is needed to achieve this, but provided the meter is not 358 asked to do it for very large numbers of flows, it has been 359 successfully implemented. 361 Inter-arrival times. The Meter knows the time that it encounters each 362 individual packet. Statistics can be kept to record the inter-arrival 363 times of the packets, which would give an indication of the jitter 364 found in the Flow. 366 Turn-around statistics. Sine the Meter knows the time that it 367 encounters each individual packet, it can produce statistics of the 368 time intervals between packets in opposite directions are observed on 369 the network. For protocols such as SNMP (where every packet elicits 370 an answering packet) this gives a good indication of turn-around 371 times. 373 Subflow analysis. Since the choice of flow endpoints is controlled 374 by the meter's rule set, it is easy to define an aggregate flow, e.g 375 "all the TCP streams between hosts A and B." Preliminary 376 implementation work suggests that - at least for this case - it 377 should be possible for the meter to maintain a table of information 378 about all the active streams. This could be used to produce at least 379 the following attributes: 381 - Number of streams, e.g. streams active for n-second intervals. 382 Determined for TCP and UDP using source-dest port number pairs. 383 - Number of TCP bytes, determined by taking difference of TCP 384 sequence numbers for each direction of the aggreagate flow. 386 IIS attributes. Work at CEFRIEL [8] has produced a traffic meter 387 with a rule set modified 'on the fly' so as to maintain a list of 388 RSVP-reserved flows. For such flows the following attributes have 389 been implemented (these quantities are defined in [9]): 391 - QoSService: Service class for the flow 392 (guaranteed, controlled load) 393 - QoSStyle: Reservation setup style 394 (wildcard filter, fixed filter, 395 shared explicit) 396 - QoSRate: [byte/s] rate for flows with 397 guaranteed service 398 - QoSSlackTerm: [microseconds] Slack Term QoS parameter 399 for flows with guaranteed service 400 - QoSTokenBucketRate: [byte/s] Token Bucket Rate QoS parameter 401 for flows with guaranteed or 402 controlled load service 404 The following are also being considered: 406 - QoSTokenBucketSize: [byte] Size of Token Bucket 408 - QoSPeakDataRate: [byte/s] Maximum rate for incoming data 410 - QoSMinPolicedUnit: [byte] IP datagrams less than this are 411 counted as being this size 412 - QoSMaxDatagramSize: [byte] Size of biggest datagram which 413 conforms to the traffic specification 415 3.6. Actions on Exceptions 417 Some users of RTFM have requested the ability to mark flows as having 418 High Watermarks. The existence of abnormal service conditions, such 419 as non-ending flow, a flow that exceeds a given limit in traffic 420 (e.g. a flow that is exhausting the capacity of the line that carries 421 it) would cause an ALERT to be sent to the Meter Reader for 422 forwarding to the Manager. Operations Support could define service 423 situations in many different environments. This is an area for 424 further discussion on Alert and Trap handling. 426 4. Extensions to the 'Basic' RTFM Meter 428 The WG has agreed that the basic RTFM Meter will not be altered by 429 the addition of the new attributes of this document. This section 430 describes the extensions needed to implement the new attributes. 432 4.1. Flow table extensions 434 The architecture of RTFM has defined the structure of flows, and this 435 draft does not change that structure. The flow table could have 436 ancillary tables called "Distribution Tables" and "Trace Tables," 437 these would contain rows of values and or actions as defined above. 438 Each entry in these tables would be marked with the number of its 439 corresponding flow in the RTFM flow table. 441 Note: The following section is for the user who is familiar with the 442 writing of rule sets for the RTFM Meter. 444 In order to identify the data in a Packet Flow Table, the attribute 445 name could be pushed into a string at the head of each row. For 446 example, if a table entry has "To Bit Rate" for a particular flow, 447 the "ToBitRate" string would be found at the head of the row. (An 448 alternative method would be to code an identification value for each 449 extended attribute and push that value into the head of the row.) See 450 section 5. for an inital set of ten extended flow attributes. 452 4.2. Specifying Distributions in RuleSets 454 At first sight it would seem neccessary to add extra features to the 455 RTFM Meter architecture to support distributions. This, however, is 456 not neccessarily the case. 458 What is actually needed is a way to specify, in a ruleset, the 459 distribution parameters. These include the number of counters, the 460 lower and upper bounds of the distribution, whether it is linear or 461 logarithmic, and any other details (e.g. the time interval for 462 short-term rate attributes). 464 Any attribute which is distribution-valued needs to be allocated a 465 RuleAttributeNumber value. These will be chosen so as to extend the 466 list already in the RTFM Meter MIB document [7]. 468 Since distribution attributes are multi-valued it does not make sense 469 to test them. This means that a PushPkt (or PushPkttoAct) action 470 must be executed to add a new value to the distribution. The old- 471 style attributes use the 'mask' field to specify which bits of the 472 value are required, but again, this is not the case for 473 distributions. Lastly, the MatchedValue ('value') field of a PushPkt 474 rule is never used. Overall, therefore, the 'mask' and 'value' 475 fields in the PushPkt rule are available to specify distribution 476 parameters. 478 Both these fields are at least six bytes long, the size of a MAC 479 address. All we have to do is specify how these bytes should be 480 used! As a starting point, the following is proposed (bytes are 481 numbered left-to-right. 483 Mask bytes: 484 1 Transform 1 = linear, 2 = logarithmic 485 2 Scale Factor Power of 10 multiplier for Limits 486 and Counts 487 3-4 Lower Limit Highest value for first bucket 488 5-6 Upper Limit Highest value for last bucket 490 Value bytes: 491 1 Buckets Number of buckets. Does not include 492 the 'overflow' bucket 493 2 Parameter-1 } Parameter use depends 494 3-4 Parameter-2 } on distribution-valued 495 5-6 Parameter-3 } attribute 497 For example, experiments with NeTraMet have used the following 498 rules: 500 FromPacketSize & 1.0.25!1500 = 60.0!0: PushPkttoAct, Next; 502 ToInterArrivalTime & 2.3.1!1800 = 60.0.0!0: PushPkttoAct, Next; 504 FromBitRate & 2.3.1!10000 = 60.5.0!0: PushPkttoAct, Next; 506 In these mask and value fields a dot indicates that the preceding 507 number is a one-byte integer, the exclamation marks indicate that the 508 preceding number is a two-byte integer, and the last number is two 509 bytes wide since this was the width of the preceding field. (Note 510 that this convention follows that for IP addresses - 130.216 means 511 130.216.0.0). 513 The first rule specifies that a distribution of packet sizes is to be 514 built. It uses an array of 60 buckets, storing values from 1 to 1500 515 bytes (i.e. linear steps of 25 bytes each bucket). Any packets with 516 size greater than 1500 will be counted in the 'overflow' bucket, 517 hence there are 61 counters for the distribution. 519 The second rule specifies an interarrival-time distribution, using a 520 logarithmic scale for an array of 60 counters (and an overflow 521 bucket) for rates from 1 ms to 1.8 s. Arrival times are measured in 522 microseconds, hence the scale factor of 3 indicates that the limits 523 are given in milliseconds. 525 The third rule specifies a bit-rate distribution, with the rate being 526 calculated every 5 seconds (parameter 1). A logarithmic array of 60 527 counters (and an overflow bucket) are used for rates from 1 kbps to 528 10 Mbps. The scale factor of 3 indicates that the limits are given 529 in thousands of bits per second (rates are measured in bps). 531 These distribution parameters will need to be stored in the meter so 532 that they are available for building the distribution. They will 533 also need to be read from the meter and saved together with the other 534 flow data. 536 4.3. Reading Distributions 538 Since RTFM flows are bi-directional, each distribution-valued 539 quantity (e.g. packet size, bit rate, etc.) will actually need two 540 sets of counters, one for packets travelling in each direction. It is 541 tempting to regard these as components of a single 'distribution,' 542 but in many cases only one of the two directions will be of interest; 543 it seems better to keep them in separate distributions. This is 544 similar to the old-style counter-valued attributes such as toOctets 545 and fromOctets. 547 A distribution should be read by a meter reader as a single, 548 structured object. The components of a distribution object are 550 - 'mask' and 'value' fields from the rule which created 551 the distribution 553 - sequence of counters ('buckets' + overflow) 555 These can be easily collected into a BER-encoded octet string, and 556 would be read and referred to as a 'distribution.' 558 5. Extensions to the Rules Table, Attribute Numbers 560 The Rules Table of "old-style" attributes will be extended for the 561 new flow types. A list of actions, and keywords, such as "ToBitRate", 562 "ToPacketSize", etc. will be developed and used to inform an RTFM 563 meter to collect a set of extended values for a particular flow (or 564 set of flows). 566 Note. An implementation suggestion. Value 65 is used for 567 'Distributions,' which has one bit set for each distribution-valued 568 attribute present for the flow, using bit 0 for attribute 66, bit 1 569 for attribute 67, etc. 571 Here are ten possible distribution-valued attributes numbered 572 according to RTFM WG consensus at the 1997 meeting in Munich: 574 ToPacketSize(66) size of PDUs in bytes (i.e. number 575 FromPacketSize(67) of bytes actually transmitted) 577 ToInterarrivalTime(68) microseconds between successive packets 578 FromInterarrivalTime(69) travelling in the same direction 580 ToTurnaroundTime(70) microseconds between successive packets 581 FromTurnaroundTime(71) travelling in opposite directions 583 ToBitRate(72) short-term flow rate in bits per second 584 FromBitRate(73) Parameter 1 = rate interval in seconds 586 ToPDURate(74) short-term flow rate in PDUs per second 587 FromPDURate(75) Parameter 1 = rate interval in seconds 589 (76 .. 97) other distributions 591 It seems reasonable to allocate a further group of numbers 592 for the IIS attributes described above - 594 QoSService(98) 595 QoSStyle(99) 596 QoSRate(100) 597 QoSSlackTerm(101) 598 QoSTokenBucketRate(102) 599 QoSTokenBucketSize(103) 600 QoSPeakDataRate(104) 601 QoSMinPolicedUnit(105) 602 QoSMaxPolicedUnit(106) 604 The following attributes have also been implemented in NetFlowMet, a 605 version of the RTFM traffic meter - 607 MeterID(112) Integer identifying the router producing 608 NetFlow data (needed when NetFlowMet takes 609 data from several routers) 610 SourceASN(113) Autonomous System Number for flow's source 611 SourcePrefix(114) CIDR width used by router for determining 612 flow's source network 613 DestASN(115) Autonomous System Number for flow's destination 614 DestPrefix(116) CIDR width used by router for determining 615 flow's destination network 617 Some of the above, e.g. SourceASN and DestASN, might sensibly be 618 allocated attribute numbers below 64, making them part of the 'base' 619 RTFM meter attributes. 621 To support use of the RTFM meter as an 'Edge Device' for implementing 622 Differentiated Services, and/or for metering traffic carried via such 623 services, two more attributes will be useful: 625 DSCodePoint(118) DS Code Point (6 bits) for packets in this flow 627 Note that since the DS Code Point is a single field within a packet's 628 IP header, it is not possible to have both Source- and Dest- 629 CodePoint attributes. Possible uses of DSCodePoint include 630 aggregating flows using the same Code Points, and separating flows 631 having the same end-point addresses but using different Code Points. 633 6. Security Considerations 635 The attributes considered in this document represent properties of 636 traffic flows; they do not present any security issues in themselves. 637 The attributes may, however, be used in measuring the behaviour of 638 traffic flows, and the collected traffic flow data could be of 639 considerable value. Suitable precautions should be taken to keep such 640 data safe. 642 7. Author's Addresses: 644 Sig Handelman 646 IBM Research Division 647 Hawthorne, NY 648 Phone: 1-914-784-7626 649 E-mail: swhandel@us.ibm.com 650 Nevil Brownlee 651 The University of Auckland 652 New Zealand 653 Phone: +64 9 373 7599 x8941 654 E-mail: n.brownlee@auckland.ac.nz 656 Greg Ruth 657 GTE Laboratories 658 Waltham, MA 659 Phone: 1 781 466 2448 660 E-mail: grr1@gte.com 662 Stephen Stibler 663 IBM Research Division 664 Hawthorne, NY 665 Phone: 1-914-784-7191 666 E-mail: stibler@us.ibm.com 668 8. References: 670 [1] Brownlee, N., Mills, C., Ruth, G., 671 "Traffic Flow Measurement: Architecture," 672 RFC 2063, The University of Auckland, January 1997. 674 [2] Wroclawski, J., "The Use of RSVP with IETF Integrated Services," 675 RFC 2210, September 1997. 677 [3] Paxson, V., Almes, G., Mahdavi, J., Mathis, M., "Framework for 678 IP Performance Metrics," RFC 2330, May 1998. 680 [4] Claffy, K., Braun, H-W, Polyzos, G., "A Parameterizable 681 Methodology for Internet Traffic Flow Profiling," IEEE Journal on 682 Selected Areas in Communications, Vol. 13, No. 8, October 1995. 684 [5] Mills, C., Hirsch, G. and Ruth, G., "Internet Accounting 685 Background," RFC 1272, Bolt Beranek and Newman Inc., 686 Meridian Technology Corporation, November 1991. 688 [6] Waldbusser, S., "Remote Network Monitoring Management Information 689 Base," RFC 1757, 1995, and RFC 2021, 1997. 691 [7] Brownlee, N: "Traffic Flow Measurement: Meter MIB", 692 RFC 2064, The University of Auckland, January 1997. 694 [8] Maiocchi, S: "NeTraMet & NeMaC for IIS Accounting: User's Guide", 695 CEFRIEL, Milan, 5 May 1998. (See also http://www.cefriel.it/ntw) 697 [9] Shenker, S., Partridge, C., Guerin, R.: "Specification of 698 Guaranteed Quality of Service," RFC 2212, 1997. 700 Expires October 1999