Patent Application
20070147402
The present invention relates generally to IP networks. More particularly, the present invention relates to Point to Point Protocol (“PPP”) links running PPP Multilink Protocol with multi-class extensions.
PPP Multilink Protocol (“MP” or “Multilink” or “Multilink Protocol”) provides a method for splitting, recombining, and sequencing datagrams across multiple logical data links (that is, interfaces). A frame is a unit of transmission in the data link layer, consisting of a frame header and a packet. An Internet Protocol (“IP”) datagram is a unit of end-to-end transmission in the IP protocol. It consists of an IP header followed by the Transmission Control Protocol (“TCP”) segment. A packet is a generic way to refer to datagrams that are either whole (that is, an IP datagram) or fragmented (that is, an IP datagram fragment). Multilink allows packets to be fragmented and the fragments to be sent over multiple PPP links to the same remote address resulting in aggregation of traffic across multiple links for extra bandwidth and redundancy.
Multilink with multi-class extensions (“Multilink-Extension”) is typically used to provide Quality of Service (“QOS”) to data traffic by using fragmentation and interleaving of packets across a multi-link interface. Thus, Multilink-Extension allows a sender to fragment packets of various priorities into multiple classes of fragments, allowing high-priority packets to be sent between fragments of lower priorities.
The present invention provides for both peers on a PPP link running Multilink-Extension to negotiate the same number of egress priority queues as well as negotiate a number of Multilink classes equal to the negotiated number of egress priority queues. Thus, both ends of the PPP link support the same number of egress priority queues and the same number of Multilink-Extension classes. This allows communication devices that have a different default number, or different maximum number, of egress priority queues to interoperate in a manner that ensures packets have the same per-hop behavior (“PHB”). The present invention is both memory efficient and processing time efficient because only the minimum number of egress priority queues necessary are instantiated.
In a first aspect, the present invention embodies a communication device comprising a number of egress priority queues determined during a Multilink-Extension negotiation. In a second aspect, the present invention embodies a PPP link comprising two peers having an equal number of egress priority queues determined during a Multilink-Extension negotiation. In a third aspect, the present invention embodies a method comprising negotiating a number of egress priority queues to instantiate during a Multilink-Extension negotiation.
As is known in the art, Multilink is based on a Link Control Protocol (“LCP”) option negotiation (“Multlink negotiation”) that permits a system to indicate to its peer that it is capable of combining multiple physical links into a “bundle.” Multilink is negotiated during the initial LCP option negotiation. A system indicates to its peer that it is willing to do Multilink by sending the Multilink option as part of the initial LCP option negotiation. This negotiation indicates three things: 1) the system offering the option is capable of combining multiple physical links into one logical link; 2) the system is capable of receiving upper layer protocol data units (“PDUs”) fragmented using the Multilink header (described herein) and reassembling the fragments back into the original PDU for processing; and 3) the system is capable of receiving PDUs of size N octets where N is specified as part of the option even if N is larger than the maximum receive unit (“MRU”) for a single physical link. Once Multilink has been successfully negotiated, the sending system is free to send PDUs encapsulated and/or fragmented with the Multilink header.
Individual fragments are the “packets” in the Multilink Protocol. Network Protocol packets are first encapsulated (but not framed) according to normal PPP procedures, and large packets are broken up into multiple segments sized appropriately for the multiple physical links. Multilink implementations do not include the Address and Control Field in the logical entity to be fragmented, even though it would otherwise be permitted by the PPP specification. A new PPP header consisting of the Multilink Identifier and the Multilink header (“MP Header”) is inserted before each section. In this manner, the first fragment of a Multilink packet in PPP will have two headers—one for the fragment, followed by the header for the packet itself. In systems implementing Multilink, there is no requirement that the segments be of equal size, or that packets must be broken up at all.
Multilink fragments are encapsulated using the protocol identifier (“PID”) 0x00-0x3d, as shown in Structure 1.
Following the protocol identifier is a 4-byte header (that is, the MP Header) containing a sequence number, and two 1-bit fields indicating that the fragment begins a packet or terminates a packet. This format is generally referred to in the art as The Long Sequence Number Fragment Format. The Beginning fragment bit, indicated as B in Structure 1, is a 1-bit field set to 1 on the first fragment derived from a PPP packet and set to 0 for all other fragments from the same PPP packet. The Ending fragment bit, indicated as E in Structure 1, is a 1-bit field set to 1 on the last fragment and set to 0 for all other fragments. A fragment may have both the Beginning and Ending fragment bits set to 1. The Sequence Field is a 24-bit sequence number that is incremented for every fragment transmitted. Between the Ending fragment bit and the sequence number is a 6-bit field that will be discussed later. In Structure 1, the bits in this 6-bit field are all set to 0.
After negotiation of an additional PPP LCP option, the 4-byte header may be optionally replaced by a 2-byte header with only a 12-bit sequence number. This format is shown in Structure 2 and is generally referred to in the art as The Short Sequence Number Fragment Format. Between the Ending fragment bit and the sequence number is a 2-bit field that will be discussed later. In Structure 2, the bits in this 2-bit field are both set to 0.
As is know in the art Multilink can be extended to multiple classes (“Multilink-Extension”). Multilink-Extension is typically used to provide Quality of Service (“QOS”) to data traffic by using fragmentation and interleaving of packets across a Multilink interface. In Multilink-Extension, when using the short sequence number format, the 2-bit field after the Ending fragment bit in the packet is used to store a class number, designating one of four possible classes. This is illustrated in Structure 3, wherein cls is the class number.
In Multilink-Extension, when using the long sequence number format, four of the bits in the 6-bit field after the Ending fragment bit in the packet is used to store the class number. This is illustrated in Structure 4, wherein cls is the 4-bit class number.
Generally, two communication devices on a PPP link, referred to herein as peers, establish Multilink-Extension via a LCP negotiation. The establishment of Multilink-Extension between two peers will be referred to herein as a Multilink-Extension negotiation. Multilink-Extension negotiations, per se, are known in the art. Communication devices of the prior art, engaging in a Multilink-Extension negotiation, establish a number of classes (referred to herein as “classes” or “PPP classes”). At the time of a prior art Multilink-Extension negotiation, each peer will have established a number of egress priority queues according to a default profile. Communication devices of the prior art each have a default number of egress priority queues that the device supports. This default number of queues will often be the maximum number of queues supportable by the communication device. The default number, and maximum number, of supportable egress priority queues on two communication devices may differ. This is frequently the situation, for example, when the two communication devices are manufactured by two different vendors. Thus, when two communication devices of the prior art engage in a Multilink-Extension negotiation, the negotiation frequently results in a Multilink-Extension wherein the two communication devices are each supporting a different number of egress priority queues. Such a prior art Multilink-Extension negotiation generally results in classes exhibiting different PHB or results in the Multilink-Extension being inoperable altogether.
In a communication device of the present invention, the number of egress priority queues to support a Multilink Extension is negotiated during the Multilink-Extension negotiation. Additionally, the number of classes negotiated during the Multilink-Extension negotiation is equal to the negotiated number of egress priority queues. Once this negotiated number of egress priority queues is determined, the queues are instantiated. In this manner, the egress priority queues are said to be instantiated dynamically. That is, the queues are instantiated during or as a result of a Multilink-Extension negotiation. During the Multilink-Extension negotiation in the present invention, the peers compare the maximum number of egress priority queues supported on each peer and then each agree to instantiate an equal number of egress priority queues, wherein the instantiated equal number of queues is equal to or less than the maximum number of queues supported by either peer. For example, two peers may enter into a Multilink-Extension negotiation with the first peer capable of supporting eight queues and the second peer capable of supporting only two queues. In accordance with the present invention, the two peers will negotiate and determine the number of queues for each peer to instantiate. In this example, the determined number of queues will be equal to or less than two. Once the number is determined, each peer will instantiate the negotiated number of queues. If, for example, the peers negotiate to instantiate two queues, then each peer will instantiate two queues. On the other hand, if the peers negotiate to instantiate only one queue, then each peer will instantiate one egress priority queue. Additionally, the two peers will each agree to establish a number of classes equal to the negotiated number of egress priority queues.
In the prior art, to the contrary, each individual peer determines the number of egress priority queues to use in a Multilink-Extension. Two peers in a Multilink-Extension negotiation frequently can have a different number of egress priority queues, which is an undesirable situation. In the present invention, on the other hand, two peers negotiate to instantiate an equal number of egress priority queues. Thus, in a Multilink-Extension of the present invention, one or both peers will typically change the number of queues that was being supported when the peers entered into the Multilink-Extension negotiation. In the example above, for example, one of the peers may have been supporting as many as eight queues when the Multilink-Extension was entered into. As a result of the Multilink-Extension negotiation, the peer supports either one or two egress priority queues.
Differentiated Services (“DS”) enhancements to the Internet Protocol are also known in the art and will not be described in detail, but will be summarized herein. Generally, a header field, called a DS field, contains six bits (“DSCP field”) that are used as a codepoint (“DS Codepoint” or “DSCP”) to select Per-Hop Behavior (“PHB”) a packet experiences at each node in a network. DS-compliant nodes in a network select PHBs by matching against the DSCP field. This can be accomplished, for example, by treating the value of the field (that is, the codepoint) as a table index which is used to select a particular packet handling mechanism, which has been implemented in that device. This selection is typically referred to as a mapping of the codepoints to PHBs. This mapping is configurable in DS-compliant nodes. Thus, a DS-compliant node supports the logical equivalent of a configurable mapping table from codepoints to PHBs.
Preferred embodiments of the present invention establish a DSCP-to-PPP class mapping in peers that negotiate Multilink-Extension. The DSCP-to-PPP class mapping can be used to make peers (that is, nodes) of the present invention DS-compliant. Examples of DSCP-to-PPP class mappings, in accordance with the present invention, are shown in Table 1. The first DSCP-to-PPP class mapping in Table 1 shows 4 PPP classes, numbered 0 to 3, resulting from a Multilink-Extension negotiation, in accordance with the present invention, using the Short Sequence Number Fragment Format With Classes as described in Structure 3 and accompanying text. Table 1 shows eight DS codepoint groups that map into the 4 PPP classes. For example, codepoints DF and CS0 are mapped to PPP class 0. Codepoints CS5, CS6, and CS7 are all mapped to PPP class 2. Similarly, Table 1 illustrates an example of mapping DSCPs into three classes (and 3 queues). Table 1 also illustrates an example of mapping DSCPs into 2 classes (and 2 queues). Thus, Table 1 illustrates how DS codepoints can be mapped into classes.
1Lower numbered queue has higher priority
2x = 1, 2, or 3
The DSCP-to-PPP class mappings shown in Table 1 also illustrate how the classes can be mapped into egress priority queues for scheduling. Each class can be assigned to an egress priority queue. For example, Table 1 (at columns 2 and 3) illustrates a class to queue mapping after two peers have negotiated to 4 queues during a Multilink-Extension negotiation in accordance with the present invention. Each of the 4 PPP classes is mapped into one of the queues shown in column 3, with class 3 mapped to the highest priority queue (that is, queue 1), class 2 mapped to queue 2, class 1 mapped to queue 3, and class 0 mapped to queue 4. In this manner, all packets of PPP class 3, will be placed in queue 1 for scheduling purposes. All packets of PPP class 2, regardless of whether codepoint CS5, CS6, or CS7, will be placed in queue 2 for scheduling purposes. All packets of PPP class 1, regardless of whether codepoint CS1, CS2, CS3, etc., will be placed into queue 2 for scheduling, and all packets of PPP class 0, regardless of whether codepoint DF or CS0, will be placed into queue 4 for scheduling. Accordingly, all packets of each PPP class will have the same PHB. Similarly, columns 4 and 5 show a PPP class to queue mapping after two peers have negotiated to 3 queues (and 3 classes) during a Multilink-Extension negotiation. The PPP class to queue mapping shown in columns 4 and 5 has class 0 mapped into queue 3. Accordingly, all packets of PPP class 0 will be placed into queue 3 for scheduling, all packets of class 1 will be placed into queue 2 for scheduling, and all packets of class 2 will be placed in queue 1 for scheduling. Columns 6 and 7 illustrate a class to queue mapping after two peers have negotiated to 2 queues (and 2 classes) during a Multilink-Extension negotiation. The class to queue mapping illustrated in columns 6 and 7 has PPP class 0 being mapped into queue 2 while PPP class 1 is mapped into queue 1.
Examples of DSCP-to-PPP class mappings, in accordance with the present invention, are also shown in Table 2. The first DSCP-to-PPP class mapping in Table 2 shows 8 PPP classes, numbered 0 to 7, resulting from a Multilink-Extension negotiation, in accordance with the present invention, using the Long Sequence Number Fragment Format With Classes as described in Structure 4 and accompanying text. Table 2 also shows 8 DS codepoint groups that map to the 8 PPP classes. Columns 1 and 2 illustrate a DSCP-to-class mapping after two peers have negotiated to 8 queues (and 8 classes) during a Multilink-Extension negotiation. Similar to Table 1, the codepoints in Table 2 are mapped into classes. However, because of the larger number of classes in Table 2, the DSCP-to-class mapping is different. In Table 1, for example, codepoints CS1, CS2, CS3, and CS4, are all mapped to class 1. In Table 2, however, codepoint CS 1 is mapped to class 1, codepoint CS2 is mapped to class 2, codepoint CS3 is mapped to class 3, and codepoint CS4 is mapped to class 4. Similarly, Table 2 shows a DSCP-to-class mapping for a situation with 6 negotiated queues (and 6 classes).
1Lower numbered queue has higher priority
2x = 1, 2, or 3
The DSCP-to-class mappings shown in Table 2 also illustrates how the classes can be mapped into egress priority queues for scheduling. For example, columns 2 and 3 illustrate a class to queue mapping after two peers have negotiated to 8 queues during a Multilink-Extension negotiation in accordance with the present invention. Each of the 8 classes is mapped into one of the queues as shown in columns 2 and 3. Class 7 is mapped to queue 1, class 6 is mapped to queue 2, and class 5 is mapped to queue 3, etc. Similarly, columns 4 and 5 in Table 2 illustrate a class to queue mapping after two peers have negotiated to 6 queues during a Multilink-Extension negotiation in accordance with the present invention.
Multilink-Extension embodiments of the present invention generally are more efficient than Multilink-Extension of the prior art. A minimum number of queues can be instantiated in each peer running Multilink-Extension in accordance with the present invention. In this manner, the peers do not use memory instantiating queues that will not be utilized. Additionally, Multilink-Extension of the present invention will not waste processing time scheduling queues that are not utilized. For example, if a communication device of the prior art is capable of supporting eight queues, but is running Multilink-Extension with only four classes, four of the eight supportable queues may not be utilized. These four unused queues would unnecessarily take up more memory than is needed and waste processing time in scheduling. Multilink-Extension in accordance with the present invention, on the other hand, would have negotiated the eight supportable queues down to four or fewer instantiated queues. Even if both peers can support up to eight queues, they will negotiate down to instantiating only four or fewer queues if the peers establish only four classes during a Multilink-Extension negotiation according to the present invention. In this manner, each peer will only use an amount of memory and processing time to support the four or fewer queues rather than the amount of memory and processing time to support eight queues.
Embodiments of the present invention comprise operable logic adapted to perform methods (for example, Multilink-Extension negotiations) of the present invention. That is, the operable logic executes the methods of the present invention. The operable logic of the present invention can be implemented as a set of computer program instructions that are stored in a computer-readable medium and executed by an embedded microprocessor system within devices made in accordance with the present invention. Embodiments of the invention may be implemented, in whole or in part, in any conventional computer programming language. For example, preferred embodiments may be implemented in a procedural programming language (for example, “C”) or an object oriented programming language (for example, “C++”). Alternative embodiments of the invention may be implemented, in whole or in part, utilizing discrete components, integrated circuitry, programmable logic used in conjunction with a programmable logic device such as a Field Programmable Gate Array (FPGA) or microprocessor, or any other means including any combination thereof. Additional embodiments of the invention may be implemented, in whole or in part, as a computer program product for use with a computer system. Such implementation may include a series of computer instructions fixed either on a tangible medium, such as a computer-readable media (for example, a diskette, CD-ROM, ROM, or fixed disk), or fixed in a computer data signal embodied in a carrier wave that is transmittable to a computer system via a modem or other interface device, such as a communications adapter connected to a network over a medium. The medium may be either a tangible medium (e.g., optical or analog communications lines) or a medium implemented with wireless techniques (e.g., microwave, infrared or other transmission techniques). The series of computer instructions embodies all or part of the functionality described herein with respect to the present invention. Those skilled in the art should appreciate that such computer instructions can be written in a number of programming languages for use with many computer architectures or operating systems. Furthermore, such instructions may be stored in any memory device, such as semiconductor, magnetic, optical or other memory devices, and may be transmitted using any communications technology, such as optical, infrared, microwave, or other transmission technologies. It is expected that such a computer program product may be distributed as a removable medium with accompanying printed or electronic documentation (for example, shrink wrapped software), preloaded with a computer system (for example, on system ROM or fixed disk), or distributed from a server or electronic bulletin board over a network (for example, the Internet or World Wide Web).
The present invention as described herein may refer to numerous associations, mappings, or correspondences. For example, in the description of DSCP-to-class mappings of the present invention DS codepoints are mapped to classes and classes are mapped to queues. As is known in the art of implementing operable logic as described above (including the art of computer programming, for example), these, and other, associations, mappings, and correspondences typically manifest themselves, for example, either as programming data structures (either separate from or embedded in the operable logic) to keep track of and manipulate the associations, mappings, and correspondences or, alternately, can manifest themselves in the procedural aspects of the operable logic.
In accordance with the present invention, novel devices and methods embodying dynamic queue instantiation have been provided. Dynamic queue instantiation can be implemented to improve the operation of prior art Multilink-Extensions. While the present invention has been shown and described herein with reference to specific embodiments thereof, it should be understood by those skilled in the art that variations, alterations, changes in form and detail, and equivalents may be made or conceived of without departing from the spirit and scope of the invention. Accordingly, the scope of the present invention should be assessed as that of the appended claims and by equivalents thereto.
