US 6,920,133 B1Grant
Techniques for introducing in-band network management packets in multi-protocol label switching networks
Issue Date:2005-07-19
•6 Claims
•7 Drawing Sheets
Abstract
A system and method is provided for introducing in-band network management packets (INMPS) in a Multi-Protocol Label Switching (MPLS) network. This system and method of the invention uses INMPs for carrying OA&M information to label switching routers (LSRs) for effectively managing and operating MPLS-based networks. This system and method of the invention also includes techniques for distinguishing INMPs from user packets in an MPLS network. The system and method of the invention further includes using a predetermined code in a shim header of an MPLS packet to determine whether an MPLS packet is an INMP or a user packet. The predetermined code may be provided in an experimental field or a time-to-live field in the shim header of the packet. Alternatively, a label may be reserved for distinguishing an INMP from a user packet.
Metadata
Assignee
- AT&T Corp.
Inventor
- Samson Boodaghians
Application Information
Application Number:US 09/589,466
Filing Date:2000-06-07
Priority Date:2000-06-07
Art Unit:2662
Patent Drawings (7 sheets)
Description
[0001] U.S. patent application Ser. No. 09/589,466, entitled “Loopback Capability for Bi-Directional Multi-Protocol Label Switching Traffic Engineered Trunks” and originally filed the same day as the present application, is hereby incorporated by reference.
Field of the Invention
[0002] The present invention relates generally to network operation, administration, and maintenance (OA&M) functions for communication networks. More particularly, the present invention relates to the use of in-band network management packets and corresponding techniques for distinguishing in-band network management packets from user packets.
Background of the Invention
[0003] A typical digital communications network has a network architecture that is based upon the Open Systems Interconnection (OSI) Reference Model for providing communication between a multiplicity of interconnected digital end systems or “nodes.” The OSI Reference Model divides networking protocols into seven layers, which, in ascending order of abstraction, are: 1) the physical layer, 2) the data-link layer, 3) the network layer, 4) the transport layer, 5) the session layer, 6) the presentation layer, and 7) the application layer.
[0004] Local area networks (LANs), i.e., short-distance communications networks, operate at layer 2 in the OSI model. Routers operate at layer 3 in the OSI model and may connect two LANs or other types of networks having different protocols. More specifically, routers at the network layer terminate local data-link layer protocols and utilize network layer addresses and data frame restructuring to communicate with each other.
[0005] Internet Protocol (IP) is a typical layer 3 routing protocol. For IP routing, a router receives a packet and determines a next hop, i.e., the next destination for the received packet in the path towards the final packet destination. Typically, each router in the path to the final destination of the packet analyzes the packet header for identifying the packet's destination address and runs a routing algorithm for determining the next hop towards the identified destination address.
[0006] Multi-Protocol Label Switching (MPLS) optimizes conventional routing techniques by assigning labels to a Forwarding Equivalent Class (FEC). A FEC is defined as a set of packets that can be handled equivalently for the purpose of forwarding and thus is suitable for binding to a label. Once a binding between a FEC and a label is done, it is not necessary for each label switching router (LSR) in a label-switched path (LSP), i.e., a path through one or more LSRs followed by packets having the same FEC, to analyze a received packet's IP header for determining the packet's destination address. Instead, LSRs make forwarding decisions based on the label attached to the packet, and consequently, packets are routed through the network faster.
[0007] MPLS is being developed for high-speed networks that, for example, are used by some Internet-service providers (ISPs). MPLS is currently being standardized by the MPLS Working Group of the Internet Engineering Task Force (IETF).
[0008] FIG. 1 illustrates an MPLS packet having data-link header 8, MPLS shim header 7, IP header 6 and payload 5. MPLS uses shim header 7 located between the data-link layer header (i.e., data-link header 8) and the network layer header (i.e., IP header 6) in the MPLS packet for integrating IP routing with label switching. MPLS uses shim header 7 encapsulation shown in FIG. 1 for transporting IP Packets. MPLS can operate on any data-link layer media (e.g., ATM, FR, PPP), but MPLS currently serves only the IP client network layer. Shim header 7 generally includes a series of label stack entries. As shown in FIG. 1 , a label stack entry contains a 20-bit label field 1, a 3-bit experimental field 2, a single bit field 3 indicating the bottom of the label stack and an 8-bit time-to-live (TTL) field 4.
[0009] Similar to conventional routing table entries, each LSR in a MPLS network may include a forwarding table having next hop label forwarding entries (NHLFEs). Each NHLFE, among other information, contains the physical interfaces or ports, the incoming label, and the outgoing label for the next hop for a received packet. A label in a label stack entry of a received packet is used as an index for retrieving a NHLFE containing the next hop for the received packet. Generally, the label from the label stack entry on the top of the label stack is used to index the NHLFEs in the LSR. After identifying the NHLFE for the next hop, the outgoing label for the next hop, which is retrieved from the identified NHLFE, is placed on top of the label stack for the packet, and the packet is transmitted to the next hop. This label switching technique is used by the LSRs for routing MPLS packets through the MPLS network.
[0010] FIG. 2 illustrates a reference topology for a typical MPLS network 45. MPLS network 45 includes a set of nodes or LSRs for performing MPLS routing and forwarding. The LSRs in MPLS network 45 include intermediate LSRs and label-edge routers, e.g., LER 1 and LER 2. The set of contiguous nodes that an MPLS packet traverses in the MPLS network is called a Label Switched Path (LSP).
[0011] MPLS network 45 includes a bi-directional traffic engineering trunk (BTT) 42 having two traffic engineering trunks with the same endpoints, i.e., LER 1 and LER 2, and opposite directions of transmission. The two traffic trunks that form BTT 42 traverse two unidirectional explicitly routed label-switched paths (ER-LSPs) between LER 1 and LER 2. An ER-LSP is a LSP defined by a management system or a single LSR that is typically the ingress LSR for that particular direction of transmission, e.g., LER 1 or LER 2. ER-LSPs are set up independent of IP shortest path routing algorithms. BTT 42 is formed of two traffic trunks traversing two ER-LSPs in opposite directions. One traffic trunk flows downstream from ingress LER 1 towards egress LER 2 on one ER-LSP. The other traffic trunk flows upstream from egress LER 2 towards ingress LER 1 on the other ER-LSP. Consequently, BTT 42 traverses a complete round-trip path between LER 1 and LER 2.
[0012] It is envisioned that ER-LSPs, which form BTTs, will be set up for carrying possibly hundreds to several thousands of individual IP flows. Hence, it is crucial to ascertain parameters of a BTT, such as connectivity, delay and other quality of service (QoS) parameters that may effect traffic flow in the BTT.
[0013] An in-band network management packet (INMP) is a packet that carries operation, administration and maintenance (OA&M) information. INMPs can be used for testing parameters of a BTT and sending OA&M commands to an LSR. U.S. patent application Ser. No. 09/589,464 describes two types of INMPs-test INMPs and command INMPs. A command INMP includes an OA&M command intended for a target LSR, which is a specific LSR targeted to receive the command INMP. Once the target LSR receives the command INMP, the LSR processes the INMP, e.g., by performing the command, and the INMP is terminated. A test INMP is used for testing parameters of a BTT. For example, a test IMLP, constructed by a LER, for testing delay and connectivity of BTT may be looped around a BTT. Each LSR in the path of the BTT receives and processes the INMP, for example, by performing a set of local operations. After processing the test INMP, the LSR receiving the test INMP transmits the test INMP to a next hop on the BTT. In the case of a loopback test, after the text INMP has been looped around the BTT and once the originating LER that originally constructed the test INMP receives the test INMP, the originating LER can ascertain delay of the BTT and whether the BTT is connected. Test INMPs may be used to test connectivity, delay and other QoS parameters.
[0014] An INMP for an MPLS network uses the same shim header 7 encapsulation that is used for an MPLS packet. An INMP, in addition to a shim header that includes a label, contains a payload, hereinafter referred to as an INMP payload. The protocol and structure of the aforementioned INMP payload is a matter for standardization and is not defined as part of this invention. However, a semantic associated with a “Switching Label Field” in the INMP payload is discussed in the following sections. Since INMPs are constructed by the service providers' LSRs and not the end-users of the network, an MPLS client layer-independent structure for the INMP messages is envisioned. Thus, INMPs can serve any client layer protocol, including IP. The advantage of this structure is that the proposed MPLS OA&M framework will be able to support any client layer including IP.
[0015] User data is conventionally carried in payload 5 of an MPLS packet. User data includes, for example, multimedia data (such as voice, fax, still and moving image, etc.), high-speed data or any user-generated information. Therefore, if an MPLS packet is an INMP, an LSR receiving the MPLS packet must have the capability to distinguish between an INMP and a user MPLS packet. For example, if a received MPLS packet is a user packet (i.e., a packet that contains user data in payload 5), the LSR receiving the packet simply determines the next hop for the packet and transmits the packet to the next hop without the need to read payload 5. However, if the packet is an INMP, the receiving LSR must read and process OA&M information in the INMP payload.
[0016] Hence, a need exists for providing LSRs with the capability to determine whether an MPLS packet is an INMP or a user packet.
Summary of the Invention
[0017] It is an aspect of the present invention to use INMPs for conveying OA&M information in an MPLS network. It is another aspect of the present invention to use different fields of the MPLS header for distinguishing INMPs from user data.
[0018] In accordance with the present invention, there is provided a system and method for distinguishing an INMP packet from a user packet. In accordance with the present invention, a predetermined 3-bit code in the EXP field of the MPLS header is used to determine whether the received packet is an INMP or a user packet. In another preferred embodiment of the present invention, a predetermined 1-bit code in the EXP field of the MPLS header is used to determine whether the received packet is an INMP or a user packet.
[0019] Alternatively, in accordance with the present invention, a predetermined code in the TTL field is used to determine whether the received packet is an INMP or a user packet.
[0020] Alternatively, in accordance with the present invention, a label (i.e., a reserved label) is designated and used to determine whether a received packet is an INMP or a user packet. Using a reserved label requires modification to conventional label switching techniques, because the reserved label is inserted in the top layer of the shim header label stack where the switching label (i.e., the label used to determine the next hop) is normally located. Thus, in accordance with, the present invention, a reserved label is inserted on top of a label stack in a shim header in a packet, and a label below the reserved label on the label stack is used to determine the next hop for the packet. Alternatively, the reserved label is inserted in a shim header of a packet, whereas the switching label is inserted in a designated field in the INMP payload.
Brief Description of the Drawings
[0021] The present invention is illustrated by way of example and not limitation in the accompanying figures in which like reference numerals indicate similar elements and in which:
[0022] FIG. 1 is a diagram illustrating a format of a packet header used in a typical MPLS network;
[0023] FIG. 2 is a schematic block diagram illustrating the topology of a typical MPLS network;
[0024] FIG. 3 is a diagram of an EXP field in an MPLS shim header;
[0025] FIG. 4 is a schematic block diagram of a label switched router of the present invention;
[0026] FIG. 5 is a diagram of a TTL field in an MPLS shim header,
[0027] FIG. 6 is a flow diagram of a preferred embodiment of the present invention that uses an INMP ID bit;
[0028] FIG. 7 is a flow diagram of a preferred embodiment of the present invention that uses an INMP ID code;
[0029] FIG. 8 is a flow diagram of a preferred embodiment of the present invention that uses the TTL field;
[0030] FIG. 9 is a flow diagram of a preferred embodiment of the present invention that uses a reserved label; and
[0031] FIG. 10 is a flow diagram of another preferred embodiment of the present invention that uses a reserved label.
Detailed Description of the Invention
1. Experimental Field (or EXP bits)
[0032] The EXP field 2 in shim header 7 of FIG. 1 has a length of three bits. FIG. 3 illustrates EXP field 2 having three bits A-C, bit A being the most significant bit and bit C being the least significant bit. Currently, the use for the EXP bits is not standardized, but there is a proposal to use the EXP bits for signaling class of service (CoS). For example, an LSR includes a plurality of queues for temporarily storing received MPLS packets that need to be routed. The MPLS packets are placed on different queues as they are ready to exit a port on the LSR. MPLS packets in the higher priority queue (e.g., the queue designated for high CoS packets) are served before the remaining packets in other queues, which contain packets of lower CoS. Alternatively, each queue can be given a share of the LSR port and line bandwidth, which is proportional to its CoS designation as compared to other queues. Since there are three bits in EXP field 2, up to eight classes can be designated.
[0033] In a preferred embodiment of the present invention, a bit in EXP field 2, hereinafter referred to as an INMP ID bit, is designated to carry coded information for differentiating INMPs from user packets. For example, bit A in EXP field 2 is designated as the INMP ID bit, and an INMP ID bit having a value of “1” identifies an INMP. One of ordinary skill in the art, however, can readily designate any one of bits A-C as the INMP ID bit, and an INMP ID bit having a value of “0” may also be used for identifying an INMP. An LSR receiving an MPLS packet with bit A in EXP field 2 having a value of “1” determines from bit A that the received MPLS packet is an INMP. After making this determination, the LSR processes the INMP to, e.g., test a parameter of the BTT or perform a command, as discussed above.
[0034] In another preferred embodiment, a predetermined 3-bit code carried in EXP field 2, hereinafter referred to as an INMP ID code, identifies an INMP. For example, “111” in EXP field 2 is designated as the INMP ID code, and an LSR receiving an MPLS packet having “111” in EXP field 2 determines that the received packet is an INMP. One of ordinary skill in the art, however, can readily designate any 3-bit code for the INMP ID code.
[0035] As discussed above, there is a proposal to use the EXP bits for designating CoS for the MPLS packet. Designating an INMP ID code limits the number of classes that may be designated from eight to seven, because one of the eight codes is allocated as an INMP ID code. Also, designating an INMP ID bit limits the number of classes that may be designated from eight to four, because one bit from EXP field 2 is allocated as an INMP ID bit. In many networks, however, up to four classes may be adequate, so allocating a code or a bit in EXP field 2 for identifying an INMP can be acceptable.
[0036] FIG. 4 is a schematic block diagram of a preferred embodiment of an LSR in MPLS network 40. The LSR shown in FIG. 4 includes ports 50-53, processing circuitry 65 and switching fabric 60. Each of ports 50-53 includes transmitting circuitry, receiving circuitry and packet assembly circuitry, as is known in the art. Ports 50-53 are connected to processing circuitry 65 through switching fabric 60. The switching fabric, for example, may be a high-speed bus and include one or more multiplexors and demultiplexors. Processing circuitry 65 includes a processor 61 and memory 62. Processor 61 may include a high-speed processor and performs forwarding/routing functions. Memory 62 may include, among other information, NHLFEs stored, for example, in a table or database for determining a next hop.
[0037] Typical operation of the LSR shown in FIG. 4 may include receiving incoming packets at ports 50-53. Packet information from the incoming packets is sent to processing circuitry 65 through switching fabric 60. Processing circuitry 65 processes the packet information to determine the next hop for an incoming packet. For example, processing circuitry 65 can identify an incoming label for each packet, and retrieve an NHLFE corresponding to the incoming label from a table stored in memory 62. The retrieved NHLFE contains the label for the next hop. Then processing circuitry 65 forwards the packet information with the new label through switching fabric 60 to the port associated with the next hop. The packet information is assembled into an outgoing packet. The outgoing packet is then transmitted to the next hop.
[0038] FIG. 6 illustrates a flow diagram of the preferred embodiment of the present invention that uses an INMP ID bit for identifying an INMP. In step 100, LER l constructs an MPLS INMP having a payload for conveying OA&M information, and LER 1 inserts the INMP ID code in shim header 7. Any LER, such as LER 2, however, may construct an MPLS INMP. In step 110, LER 1 transmits an MPLS packet on a traffic trunk, such as BTT 42, to LSR 1, i.e., the next hop on BTT 42 downstream. In step 120, LSR 1 receives the MPLS packet on, for example, port 50. The packet information is forwarded to processing circuitry 65. In step 130, processing circuitry 65 reads the MPLS packet header, including EXP field 2 in shim header 7. In step 140, processing circuitry 65 identifies the INMP ID bit from EXP field 2 and compares the INMP ID bit with a predetermined code, such as “0” or “1”. If the INMP ID bit equals the predetermined code, the processing circuitry identifies the packet as an INMP. If the INMP ID bit equals the predetermined code, processing circuitry 65 processes the INMP. In step 100, the MPLS packet is constructed as an INMP, and thus, in step 150, that INMP is processed. Processing the INMP, for example, includes determining whether the INMP is a test INMP or a command INMP. If the INMP is a test INMP, processing the INMP, for example, includes testing a parameter of the BTT and transmitting the INMP to the next hop. If the INMP is a command INMP, processing the INMP, for example, includes performing the command and terminating the INMP if the command is designated for that LSR. If the INMP ID bit is not equal to the predetermined code, in step 160, processing circuitry 65 determines a next hop, using a NHLFE, for the received packet and the MPLS packet is label-switched to LSR 2, i.e., the next hop on BTT 42.
[0039] FIG. 7 illustrates a flow diagram of the preferred embodiment of the present invention that uses an INMP ID code for identifying an INMP. The steps are substantially the same as those shown in FIG. 6 for the preferred embodiment using an INMP ID bit, except the INMP ID code is compared to a predetermined code during step 140.
2. Time-to-Live Field
[0040] Currently, as shown in FIG. 1 , the one octet TTL field 4 in MPLS shim header 7 is used in the same manner as the TTL field in IP header 6. That is, the IP TTL field is copied into the MPLS header TTL field upon a packet's entry into MPLS network 45 of FIG. 2. The MPLS header TTL field is then decremented by one as the MPLS packet traverses each LSR towards its final destination. If the TTL field becomes zero, the packet is discarded since that is a strong indication that the MPLS packet is routed in an endless routing loop among LSRs in MPLS network 45.
[0041] In another preferred embodiment of the present invention, TTL field 4 of FIG. 1 can be used to code information to distinguish INMPs from user packets. In this preferred embodiment, a single bit from TTL field 4, hereinafter referred to as a TTL ID bit, is designated as a flag to differentiate an INMP from a user packet, similarly to an INMP ID bit in EXP field 2. TTL field 4 is shown in FIG. 5 with A1 as the most significant bit. Bit A1 in TTL field 4 is designated as the TTL ID bit, and the TTL ID bit having a value of “1” identifies an INMP. One of ordinary skill in the art, however, can readily designate any one of bits A1-H1 as the TTL ID bit, and a TTL ID bit having a value of “0” may also be used for identifying an INMP. An LSR receiving an MPLS packet with bit A1 in TTL field 4 having a value of “1” determines from bit A1 that the received packet is an INMP. After determining that the packet is an INMP, the LSR processes the INMP.
[0042] Designating a bit in the TTL field for distinguishing INMPs from user data limits the number of hops a packet can traverse to 128 hops (i.e., 2 to the power of 7) before it is discarded. However, 128 hops is a more than adequate threshold for discarding packets routed in an endless loop. Also, similar to the INMP ID code, an 8-bit TTL code may be designated and used to determine whether a received packet is an INMP or a user packet.
[0043] For Hop-by-Hop MPLS, where the LSPs are formed using IP routing algorithms, the TTL field is necessary, because the potential for routing loops exists. For explicitly routed LSPs (ER-LSPs) the routes are calculated independent of IP routing algorithms, and thus the TTL field may not be necessary since the potential for routing loops can be eliminated. However, the network will still have to keep track of the TTL field as if the packet has traversed an IP network. In other words, when the packet emerges from the MPLS domain and into another MPLS or IP domain, it must contain the correct TTL value. The situation where the TTL field is not available to decrement occurs in ATM and FR MPLS networks, because ATM and FR layer 2 headers do not include a TTL field. In such cases, the Label Distribution Protocol (LDP) utilizes the Hop-Count variable to pre-calculate the hop count associated with an LSP. At the egress LER, this Hop-Count is then subtracted from the original TTL value of the packet (i.e., the TTL value of the packet upon entry into the MPLS domain). When the INMP ID code method is used to identify an INMP, the TTL field in not available to decrement. Therefore, the aforementioned Hop-Count method can be used for processing the TTL field upon exiting the MPLS domain.
[0044] FIG. 8 illustrates a flow diagram of the preferred embodiment of the present invention that uses a TTL ID bit for identifying an INMP. In step 300, LER 1 constructs an MPLS INMP having a payload for conveying OA&M information and inserts the TTL ID bit in shim header 7. Any LER, however, may construct the MPLS INMP, such as LER 2. In step 310, LER 1 transmits an MPLS packet on a traffic trunk, such as BTT 42 of FIG. 4 to LSR 1, i.e., the next hop on BTT 42 downstream. In step 320, LSR 1 receives the MPLS packet on, for example, port 50 of FIG. 4. The packet information is forwarded to processing circuitry 65 of FIG. 4. In step 330, processing circuitry 65 reads the MPLS packet header, including TTL field 4 in shim header 7. In step 340, processing circuitry 65 identifies the TTL ID bit from TTL field 4 and compares the TTL ID bit with a predetermined code, such as “0” or “1”. If the TTL ID bit equals the predetermined code, the processing circuitry identifies the packet as an INMP. If the TTL ID bit equals the predetermined code, in step 350, processing circuitry 65 processes the INMP. In step 300, the constructed MPLS packet is an INMP, and thus, in step 350, the INMP is processed. Processing the INMP, for example, includes determining whether the INMP is a test INMP or a command INMP. If the INMP is a test INMP, processing the INMP, for example, includes testing a parameter of the BTT and transmitting the INMP to the next hop. If the INMP is a command INMP, processing the INMP, for example, includes performing the command and terminating the INMP if the command is designated for that LSR. If the INMP ID bit is not equal to the predetermined code, in step 360, processing circuitry 65 determines that the packet is a user packet. Then, processing circuitry 65 determines the next hop, using a NHLFE, and the MPLS packet is label-switched to its next hop.
3. Reserved Label
[0045] An MPLS packet is transported on an LSP using the existing label-switching technique described above. For example, an LSR in network 45 receiving an MPLS packet determines the next hop for the received packet using a label from the top of the label stack. A label, however, may be reserved for identifying an INMP (i.e., distinguishing an INMP from a user packet). The reserved label is a specific label that identifies an INMP, and thus the reserved label can be used to distinguish an INMP from a user packet.
[0046] Using the reserved label requires some enhancements to the existing label-switching methods to support transport of INMPs. In a preferred embodiment of the present invention, a label stack entry including a reserved label is inserted on top of the label stack, and the switching label is inserted in a label stack entry directly below the reserved label. Once an LSR identifies the reserved label, the LSR maintains the reserved label on top of the label stack, but the LSR switches labels using the switching label in the label stack entry immediately below the reserved label. Therefore, the LSR determines the next hop for the packet using a label directly below the label (i.e., the reserved label) on top of the label stack instead of using the label on top of the label stack.
[0047] FIG. 9 illustrates the preferred embodiment of the present invention that maintains the reserved label on top of the label stack. In step 400, ingress LER 1 constructs an INMP having an INMP payload for conveying OA&M information. LER 1 then inserts the reserved label on top of the label stack in shim header 7. Note that the label normally used for switching (i.e., the switching label) now resides immediately below the aforementioned reserved label. In step 410, LER 1 transmits the INMP on BTT 42 to LSR 1, i.e., the next hop downstream on BTT 42, using the switching label. Any LER, however, may construct the INMP, and the INMP may be transmitted on a unidirectional traffic trunk as well as a bi-directional traffic trunk. In step 420, LSR 1 receives the INMP on, for example, port 50. The packet information is forwarded to processing circuitry 65. In step 430, processing circuitry 65 reads the label from the top of the label stack in shim header 7. In step 440, processing circuitry 65 determines whether the label is the reserved label. If the label read from the top of the label stack is the reserved label, the processing circuitry identifies the received packet as an INMP. In step 450, if the label read from the top of the label stack is the reserved label, processing circuitry 65 processes the INMP. In step 400, the constructed MPLS packet is an INMP, and thus, in step 450, the INMP is processed. Processing the INMP, for example, may include determining whether the INMP is a test INMP or a command INMP and transmitting the INMP to a next hop if the INMP is a test INMP or performing the command and terminating the INMP if the INMP is a command INMP. In step 450, if the INMP is not terminated, processing circuitry 65 determines the next hop for the received packet using a label in a label stack entry directly below the reserved label, which is on top of the label stack. The INMP is then transmitted to the identified next hop using the label directly below the reserved label. In step 460, if the label on top of the label stack is not the reserved label, processing circuitry 65 determines the next hop for the received packet using the label from the top of the label stack.
[0048] In another preferred embodiment of the present invention, a reserved label is carried in a label stack entry in shim header 7, and a field called the “switching label field” in the INMP payload is designated for carrying the switching label. In this embodiment, like the previously described embodiment, the reserved_label is carried in a label stack entry in shim header 7. However, in this preferred embodiment, the LSR receiving the INMP determines the next hop using a switching label carried in a field in the INMP payload of the received INMP. For example, the INMP payload may be partitioned into a plurality of fields, and one of the fields may be designated for carrying the switching label. Currently, the protocol and semantic structure of an INMP payload is not determined, and one of ordinary skill in the art may designate a field for carrying the switching label when the protocol and complete semantic structure of the INMP payload is standardized.
[0049] FIG. 10 illustrates the preferred embodiment of the present invention that carries the switching label in a field in an INMP payload without regard to its position in the stack in contrast to FIG. 9. In step 500, ingress LER 1 constructs an INMP having the reserved label in a label stack entry in shim header 7. In step 510, LER 1 transmits the INMP on BTT 42 to LSR 1, i.e., the next hop on BTT 42 downstream. Any LER, however, may construct the INMP and the INMP may be transmitted on a unidirectional as well as a bi-directional traffic trunk. In step 520, LSR 1 receives the INMP on, for example, port 50 of FIG. 4. The packet information is forwarded to processing circuitry 65 of FIG. 4. In step 530, processing circuitry 65 reads the label from shim header 7. In step 540, processing circuitry 65 determines whether the switching label is the reserved label. If the label read from a shim header 7 is the reserved label, processing circuitry 65 identifies the packet as an INMP and processes the INMP during step 550. In step 500, the constructed MPLS packet is an INMP, and thus, in step 550, the INMP is processed. Processing the INMP, for example, includes determining whether the INMP is a test INMP or a command INMP. If the INMP is a test INMP, processing the INMP, for example, includes testing a parameter of the BTT and transmitting the INMP to the next hop. If the INMP is a command INMP, processing the INMP, for example, includes performing the command and terminating the INMP if the command is designated for that LSR. If the INMP is not terminated, during step 550 processing circuitry 65 determines the next hop for the received packet using a switching label in a field in the INMP payload and the reserved label is maintained in shim header 7. Processing circuitry 65 can either switch the switching label in this field with the outgoing label identified by the retrieved NHFLE, or processing circuitry 65 can append the outgoing label to the previously appended switching label(s) in the designated field in the INMP payload. With the latter approach, a label trace of the INMP is maintained in the designated field of the INMP payload. In step 560, if the label in shim header 7 is not the reserved label, processing circuitry 65 determines the next hop for the received packet using the label on top of the label stack in shim header 7.
[0050] The present invention is applicable to Internet backbones and enterprise networks which use MPLS as a transport mechanism. In addition, some aspects of the present invention can be implemented for ATM, Frame Relay, and optical networks utilizing label-switching techniques.
[0051] What has been described are the preferred embodiments of the present invention. It will be apparent, however, to those skilled in the art that it is possible to embody the invention in specific forms other than those disclosed in the preferred embodiments described above. This may be done without departing from the spirit of the invention, and the preferred embodiments are merely illustrative and should not be considered restrictive in any way. The scope of the invention is given by the appended claims, rather than the preceding description.
Claims
1. A method of introducing in-band network management packets in a network comprising steps of:
constructing a packet including a header:
inserting a predetermined code in a field in the header; and
determining whether the packet includes an in-band network management packet, or a user packet using the predetermined code,
wherein
the constructed packet is a multi-protocol label switching packet, and
the field for inserting the predetermined code is a time-to-live field.
2. The method of claim 1, wherein the predetermined code is a one-bit code.
3. A method of introducing in-band network management packets in a multi-protocol label switching network, comprising a step of:
determining whether a packet is an in-band network management packet or a user packet by predetermined code to distinguish an in-band network management packet from a user packet; where
the packet includes a shim header and the predetermined code is inserted in an experimental field located in the shim header.
4. A method of introducing in-band network management packets in a multi-protocol label switching network, comprising a step of:
determining whether a packet is an in-band network management packet or a user packet by predetermined code to distinguish an in-band network management packet from a user packet; where
the packet includes a shim header and the predetermined code is inserted in a time-to-live field located in the shim header.
5. A network comprising:
an originating router constructing an in-band network management packet; and
a receiving router that receives a packet and determines whether the packet is an in-band network management packet or a user packet;
wherein
the originating router inserts a predetermined code in a header in the in-band network management packet, and the predetermined code identifies an in-band network management packet, and
the header includes a shim header, and the predetermined code is inserted in a time-to-live field in the shim header.
6. A router comprising:
reception circuitry that receives an incoming packet; and
processing circuitry that identifies a predetermined code and determines whether the incoming packet is an in-band network management packet or a user packet using the predetermined code, wherein the processing circuit identifies the predetermined code from a time-to-live field in a shim header of the received packet.
Patent Citations (5)
Non-Patent Literature (28)
- Comer “Internetworking with TCP/IP vol. I: Principles, Protocols, and Architecture” Third Edition, Prentice Hall, 1995, pp. 92, 93, 99, 100, 123-138, and 162.Examiner
- Rosen, MPLS Label Stack Encoding, Jan. 2001 (update to Apr. 1999), IETF RFC 3032, pp. 1-17.Examiner
- Neil Harrison—Discussion Paper—“MPLS OAM Functions and PDU Structures”, Oct. 1999.Applicant
- Neil Harrison—Discussion Paper—“OAM Principles for Layer Networks”.Applicant
- Bonica et al.—“ICMP Extensions for MultiProtocol Label Switching”, Dec. 1999.Applicant
- Callon et al—“A Framework for Multiprotocol Label Switching”, Sep. 1999.Applicant
- Rosin et al.—“Multiprotocol Label Switching Architecture”, Aug. 1999.Applicant
- LaFaucheur et al.—“MPLS Support of Differentiated Services”, Mar. 2000.Applicant
- Brin et al.—“Per Hop Behavior Identification Codes”, Oct. 1999.Applicant
- Krishnaswamy et al.—“MPLS Control Plane for Switched Optical Networks”, Feb. 2000.Applicant
- Suzuki—“The Assignment of the Information Field and Protocol Identifier in the Q.2941 Generic Identifier and Q.2957 User-to-User Signaling for the Internet Protocol”, Jan. 2000.Applicant
- Yahong—“A Probe Method in MPLS Loop Avoidance”, Dec. 1999.Applicant
- Ohba et al.—“MPLS Loop Prevention Mechanism”, Oct. 1999.Applicant
- Gibson—“The Management of MLPS”, Mar. 2000.Applicant
- Andersson et al.—“MPLS Capability Set”, Oct. 1999.Applicant
- Fedyk et al.—“Metrics and Resource Classses for Traffic Engineering”, Oct. 1999.Applicant
- Nagami, et al.—“VCID Notification Over ATM Link for LDP”, Jul. 1999.Applicant
- Ashwood-Smith et al.—“Improving Topology Data Base Accuracy With LSP Feedback Via CR-LDP”, Feb. 2000.Applicant
- Wright et al.—“CR-LDP Extensions for Interworking With RSVP-TE”, Mar. 2000.Applicant
- Andersson et al., “LDP Specification”, Oct. 1999.Applicant
- Awduche et al., “RSVP-TE: Extensions to RSVP for LSP-Tunnels”, Feb. 2000.Applicant
- Awduche et al., “Applicability Statement for Extensions to RSVP for LSP-Tunnels”, Apr. 2000.Applicant
- Boscher et al., “LDP State Machine”, Jan. 2000.Applicant
- Lang et al., “Extensions to RSVP for Optical Networking”, Sep. 2000.Applicant
- Lang et al., “Link Management Protocol (LMP)”, Sep. 2000.Applicant
- Martini et al., “Transport of Layer 2 Frames over MPLS”, Dec. 1999.Applicant
- Shew, “Fast Restoration of MPLS Label Switched Paths”, Oct. 1999.Applicant
- T. Theimer “Requirements for OAM Functionality in MPLS”, Oct. 1999.Applicant