Vlan tunneling protocol

Abstract

A virtual local area network (VLAN) tunneling system includes an ingress edge switching node that adds VLAN encapsulation information to a packet even if the egress port is configured to act as an untagged 802.1Q port. The packet is tunneled via a label-switched path (LSP) according to a multiprotocol label switching (MPLS) protocol. Label values are used for identifying a next switching node in the LSP to which the packet is to be transmitted. At a penultimate switching node in the LSP, a current label value is replaced with a label value reserved for packets originating from a port associated with a VLAN. An egress switching node in the LSP receives the packet with the reserved label value and recognizes that VLAN information is embedded in the packet. The egress switching node extracts the embedded VLAN information as well as the original source and destination addresses, and processes the packet for transmitting to a final destination.

Description

FIELD OF THE INVENTION

[0002] This invention relates generally to virtual bridged local area networks, and more particularly, to tunneling packets in such virtual bridged local area networks without loss of VLAN assignment information.

BACKGROUND OF THE INVENTION

[0003] Recent vintage switching nodes often support virtual bridged local area networks (LANs). In virtual bridged LANs, data units (e.g. packets or frames) are classified into virtual LANs (VLANs) in order to differentiate service within a bridged LAN. IEEE Draft standard 802. 1Q entitled “IEEE Standard for Local and Metropolitan Area Networks: Virtual Bridge Local Area Networks,” 1998, and IEEE Draft Standard 802.1V entitled “Draft Standard for Supplement to IEEE 802.1Q: IEEE Standard for Local and Metropolitan Area Networks: Virtual Bridge Local Area Networks,” 2000, the contents of which are hereby incorporated by reference, provide standard VLAN classification rules.

[0004] Standard 802.1Q provides VLAN tagging rules for optionally adding a tag header including the assigned VLAN identifier to the unit prior to transmitting the data unit on an egress port. Tags are applied or not depending on the VLAN of the data unit. The egress port tags the data unit unless the VLAN of the data unit belongs to an untagged set of VLANs. In this instance, the data unit is transmitted without the VLAN tag header via an untagged egress port.

[0005] Recent vintage switching nodes also often support multiprotocol label switching (MPLS). The MPLS protocol is described in detail in “Multiprotocol Label Switching Architecture,” E. Rosen et al., Internet Engineering Task Force Request for Comment 3031, January 2001 (hereinafter referred to as RFC 3031), the content of which is incorporated herein by reference. The MPLS protocol provides a connection-oriented service that enables tunneling across a wide area network. Unlike the hop-by-hop, on-demand forwarding of conventional Layer 2 (e.g. bridging) and Layer 3 (e.g. routing) protocols, the MPLS protocol provides a common protocol for end-to-end switching over heterogeneous switching nodes, referred to as label switch routers (LSRs), on pre-configured label switched paths (LSPs). A label switched path is a path through an MPLS network so that when a label is applied, traffic transits multiple routers in the LSP.

[0006] One use of the MPLS protocol is interconnection of bridged LANs over wide area networks. A primary goal of such MPLS implementations is seamless communication. That is, communication with a target host on a remote LAN (e.g. across the wide area network) should, to the extent possible, resemble communication with a target host on a local LAN. According to the MPLS protocol, this is preferably accomplished via labels attached to packets to be forwarded. When a packet is forwarded to its next hop, the assigned label is sent along with it. Analysis of the packet's network layer header is preferably done only once, and not repeated in subsequent hops. Rather, the label is used as an index into a table which specifies the next hop, and a new label. The old label is replaced with the new label, and the packet is forwarded to its next hop.

[0007] When the interconnected LANs are virtual bridged LANs, there are some potential obstacles to such seamless communication. For instance, if an egress port for an LSP is configured as an untagged 802.1Q egress port, the egress node for the LSP may not be able to readily determine the VLAN assignment for a data unit received over the LSP in order to properly process the underlying Layer 2 data unit. Even if the egress port is a tagged 802.1Q egress port, if the egress node belongs to a VLAN different from the VLAN of an ingress node, the original VLAN assignment will likely be lost. Moreover, if the egress node for the LSP supports different types of MPLS traffic, including non-802.1Q traffic, the egress node may not be able to readily differentiate traffic received over the 802.1Q LSP from other MPLS traffic in order to properly recover and process the underlying Layer 2 data unit.

[0008] Accordingly, there is a need for a system and method that allows application of the MLPS protocol virtual bridged LANs without loss of VLAN assignment information. Such a system and method should further process different types of MPLS traffic while allowing the underlying Layer 2 data unit to be properly recovered and processed.

SUMMARY OF THE INVENTION

[0009] The present invention provides a VLAN tunneling protocol that improves seamless interconnection of 802.1Q bridged LANs over wide area network (WANs). In one embodiment of the invention, a virtual local area network (VLAN) tunneling system includes a first switching node including an ingress port and an egress port. The ingress port is associated with a VLAN and receives a data packet. The egress port is configured as an untagged port. As used herein, an untagged port is a port that, under VLAN classification rules, would typically not include a VLAN tag header to an outgoing packet. The egress port however adds VLAN information to the data packet and transmits the data packet on a label-switched path.

[0010] The VLAN tunneling system further includes a second switching node that receives the data packet transmitted from the first switching node via the label-switched path. The second switching node identifies the data unit as a virtual bridged LAN data unit, retrieves the added VLAN information from the data packet, and transmits the data packet to a final destination based on the retrieved VLAN information.

[0011] In a further embodiment of the invention, the VLAN tunneling system includes a third switching node in the label-switched path which is configured to add to the packet a label value reserved for packets originating from a port associated with a VLAN.

[0012] In another embodiment of the invention, the third switching node receives the packet with the reserved label value and determines that the packet includes VLAN information associated with an originating port based on the reserved label value.

[0013] In yet another embodiment of the invention, A virtual local area network (VLAN) tunneling system includes a first switching node having a plurality of ports at least one of which has a VLAN associated therewith, and a second switching node. A data unit for transmission from said first switching node to said second switching node over said at least one port is checked for VLAN assignment prior to transmission. The data unit is transmitted from said first switching node to said second switching node on a label-switched path or not depending on a result of said check.

DESCRIPTION OF THE DRAWINGS

[0014] These and other features, aspects and advantages of the present invention will be more fully understood when considered with respect to the following detailed description, appended claims, and accompanying drawings where:

[0015] FIG. 1 is a schematic block diagram of a VLAN tunneling system according to one embodiment of the invention;

[0016] FIG. 2 is a schematic block diagram of a VLAN tunneling system where a source host transmits a data packet to a target host;

[0017] FIG. 3 is a more detailed block diagram of edge and core switching nodes in a label-switched path according to one embodiment of the invention;

[0018] FIG. 4 is as schematic diagram of a packet transmitted by the source host of FIG. 2 according to one embodiment of the invention;

[0019] FIG. 5 is a schematic diagram of the packet of FIG. 4 after being processed by an egress queue manager of an ingress edge switching node according to one embodiment of the invention;

[0020] FIG. 6A is a schematic diagram of the packet of FIG. 5 after being processed by an egress flow resolution logic of an ingress edge switching node according to one embodiment of the invention;

[0021] FIG. 6B is a schematic diagram of the packet of FIG. 5 after being processed by an egress flow resolution logic of an ingress edge switching node according to an alternative embodiment of the invention;

[0022] FIG. 7 is a schematic diagram of the packet of FIG. 6A or 6B after being processed by a penultimate core switching node in a label-switched path according to one embodiment of the invention; and

[0023] FIG. 8 is a flow diagram of a process for transmitting a data packet from the source host to the target host of FIG. 2 according to one embodiment of the invention.

DETAILED DESCRIPTION OF THE SPECIFIC EMBODIMENTS

[0024] FIG. 1 is a schematic block diagram of a VLAN tunneling system according to one embodiment of the invention. The system preferably improves seamless interconnection of 802.1Q bridged LANs over wide area networks without loss of VLAN assignment information or other underlying Layer 2 data units.

[0025] Preferably, the VLAN tunneling system includes hosts 10, 28 seamlessly receiving and transmitting data packets, also referred to as frames, over a label switched path (LSP) 30 according to a multiprotocol label switching (MPLS) protocol. The LSP is preferably formed from edge switching node 14 to edge switching node 24 over a wide area network (WAN) 16 via one or more core switching nodes 18, 20, 22.

[0026] The hosts 10, 28 are respectively connected to either edge switching node 14 or 24 over a local area network (LAN) 12 or 26 communication media such as, for example, Ethernet or Token Ring. The hosts 10, 28 are preferably network end-stations such as, for example, personal computers, workstations, servers, or other end user devices.

[0027] The edge and core switching nodes 14-24 are preferably gateway devices such as, for example, switches, routers, and the like, having network interfaces for forwarding packetized communications originated by the hosts 10, 28. The edge and core switching nodes preferably support the MPLS protocol as set forth in RFC 3031. The edge and core switching nodes are also commonly referred to as label switch routers (LSRs). Edge switching nodes 14, 24 are each commonly referred to as an ingress LSR or egress LSR depending on the direction of the flow of traffic.

[0028] The LANs 12, 26 preferably include one or more VLANs which are logical subnetworks within a bridged LAN that differentiate service based on policies rather than physical location. Each VLAN preferably includes a plurality of network devices, such as, for example, servers, workstations, and PCs, together forming a logical work group within a larger backbone network. In the embodiment illustrated in FIG. 1, hosts 10 and 28 belong to the same VLAN.

[0029] The hosts 10, 28, LANs 12, 26, and edge and core switching nodes 14-24 may be interconnected via cables or other transmission media, and may support various data communication protocols such as, for example, Ethernet, Internet Protocol (IP), and Asynchronous Transfer Mode (ATM).

[0030] FIG. 2 is a schematic block diagram of a VLAN tunneling system where a source host 40 transmits a data packet to a target host 56. The source and target hosts 40, 56 are similar to the hosts 10 and 28 of FIG. 1. Both the source host 40 and the target host 56 are preferably associated with a first VLAN 42, 54.

[0031] The data packet travels from the source host to the target host over an LSP 58. The LSP 58 preferably begins with an ingress edge switching node 44 and ends with an egress edge switching node 52, with multiple core switching nodes 46, 48, 50 in-between. The edge and core switching nodes 44-52 are similar to the edge and core switching nodes 14-24 of FIG. 1. According to the embodiment illustrated in FIG. 2, the penultimate core switching node 50 is preferably a node in the LSP 58 which is coupled to the egress edge switching node 52 via a second VLAN 51.

[0032] In general terms, the source host 40 transmits a data packet to the ingress edge switching node 44 through a port on the first VLAN 42. The ingress edge switching node receives the packet and applies a VLAN encapsulation header as set forth in the 802.1Q Standard. The VLAN encapsulation header preferably includes a VLAN ID of the first VLAN, namely a VLAN ID of “1.” Preferably, the VLAN encapsulation header is added even if the egress port forwarding the packet is an untagged port.

[0033] The ingress edge switching node 44 further applies an MPLS header to the packet as set forth in RFC 3031. The MPLS header is used to tunnel the packet from the ingress edge switching node 44 to the egress edge switching node 52 via the LSP 58 in a seamless manner. The ingress edge switching node 44 further applies a Layer 2 (Data link/MAC layer) delivery header associated with the first hop, that is, core switching node 46, to allow proper delivery to the first hop.

[0034] As the data packet traverses the LSP tunnel, MPLS label swapping occurs in a conventional fashion as set forth in RFC 3031. At the penultimate core switching node 50, the node replaces the MPLS label with a new label value reserved for virtual bridged LAN data units. The reserved label value preferably informs the egress edge switching node 52 that the packet is a tunneled Ethernet frame, and that VLAN information is embedded within it.

[0035] The egress edge switching node 52 receives the tunneled packet, for example, through a port on the second VLAN 51. The egress edge switching node 52 preferably extracts the embedded VLAN ID from the packet and performs standard processing according to the 802.1Q Standard. The VLAN ID used, however, is the extracted VLAN ID embedded in the packet, that is, a VLAN ID of “1,” instead of the VLAN ID associated with the port on which the packet was received, that is, a VLAN ID of “2.” The originally assigned VLAN information is therefore maintained while traversing the LSP.

[0036] FIG. 3 is a more detailed block diagram of the edge and core switching nodes 44-52 according to one embodiment of the invention. The nodes 44-52 each preferably include an iyngress flow resolution logic (FRL) 60, egress queue manager (EQM) 62, and egress FRL 64 for each switching interface, such as a port, of the node. The ingress and egress FRLs preferably classify and route incoming flows of packets. The EQM 62 preferably manages queues of packets for transmission out of the node's ports.

[0037] An original packet transmitted by the source host 40 is received by an ingress port (not shown) of the ingress edge switching node 44 and processed by its ingress FRL 60 in a conventional fashion. For instance, the ingress FRL 60 may check a destination source address for source learning and perform a destination address lookup and filtering. The ingress FRL 60 preferably selects an egress port (not shown) and transmits the packet to the egress port for forwarding the packet.

[0038] The egress port receives the packet and invokes its EQM 62 to enqueue the packet for transmission out of the port. When ready to be transmitted, the EQM 62 dequeues the packet and applies a VLAN tag to the packet, preferably according the 802.1Q Standard. The VLAN tag is applied even if the egress port is an untagged port. For an untagged port, however, an egress frame directed for the port's own VLAN is preferably not tagged.

[0039] The dequeued packet is transmitted by the EQM 62 to the egress FRL 64 which encapsulates it into an MPLS packet, with a source address set to the ingress edge switching node's address and a destination address set to the address corresponding to the first hop, that is, core switching node 46. The egress FRL 64 further attaches other header data as necessary and delivers the packet to the first hop.

[0040] As the packet traverses the LSP 58, it is received by an ingress port of each core switching node and processed by its ingress FRL 60. The ingress FRL identifies the MPLS packet by its Ethertype (Etype) protocol identification. The EQM 62 in the egress port of each core switching node replaces or removes the Ethernet header based on the link technology joining one core switching node to another. The egress FRL 64 in each core switching node performs appropriate MPLS label switching to appropriately follow the LSP 58. During the MPLS label switching process at the penultimate core switching node 50, however, the associated egress FRL 64 replaces the MPLS label with the reserved virtual bridged LAN data unit label.

[0041] After traversing the LSP 58, the packet is received by an ingress port (not shown) of the egress edge switching node 52 and processed by its ingress FRL 60. The ingress FRL 60 preferably identifies the reserved label inserted at the penultimate core switching node 50 and thus recognizes that VLAN information is embedded in the packet. The ingress FRL extracts the embedded VLAN information as well as the original source and destination addresses from the packet. The packet is then processed using the recovered values for transmitting to a final destination.

[0042] FIG. 4 is as schematic diagram of a packet transmitted by the source host 40 according to one embodiment of the invention. The packet preferably includes a destination address 70 of the target host 56, source address 72 of the source host 72, a protocol identifier 74, and payload data 76. If the packet is an Ethernet frame, the source and destination addresses are preferably Layer 2/MAC addresses. The protocol identifier 74 is preferably an Ethernet protocol.

[0043] FIG. 5 is a schematic diagram of the packet of FIG. 4 after being processed by the EQM 82 of the ingress edge switching node 44 according to one embodiment of the invention. The EQM 82 preferably adds a VLAN tag to the packet including a VLAN protocol type 80 and VLAN ID of the VLAN to which the source host 40 belongs even if the transmitting egress port is an untagged port. The VLAN protocol type 80 preferably identifies the 802.1Q Standard or any other VLAN classification protocol, as the protocol used for classifying the packet into a VLAN.

[0044] FIG. 6A is a schematic diagram of the packet of FIG. 5 after being processed by the egress FRL 64 of the ingress edge switching node 44 according to one embodiment of the invention. For purposes of the example illustrated in FIG. 6A, it is assumed that the egress port transmitting the packet to the first hop is an untagged port. The egress FRL 64 preferably encapsulates the packet illustrated in FIG. 5 into an MPLS packet. In this regard, the egress FRL 64 adds MPLS header information 90 including a destination address 92 that corresponds to the address of the first hop, that is, core switching node 46, and a source address 94 that corresponds to the source address of the ingress edge switching node 44. A protocol type 96 further identifies that the MPLS protocol is used for transmitting the packet. The MPLS header information also includes a label 98 used to identify the tunnel to the first hop. Preferably the entire packet illustrated in FIG. 5 that is received from the EQM 62 is included as the MPLS payload data 100.

[0045] FIG. 6B is a schematic diagram of the packet of FIG. 5 after being processed by the egress FRL 64 of the ingress edge switching node 44 according to another embodiment of the invention. For purposes of the example illustrated in FIG. 6B, it is assumed that the egress port for transmitting the packet to the first hop is a tagged port. As in the packet of FIG. 6B, the packet includes MPLS header information including the destination address 92 of the first hop, source address 94 of the ingress edge switching node 44, MPLS protocol type 96, and first hop label 98. In addition, because the egress port is a tagged port, the egress FRL 64 adds to the packet a VLAN header data including a VLAN classification protocol 110 and VLAN ID 112 of the VLAN to which the ingress edge switching node belongs for providing correct Layer 2 connectivity to the first hop.

[0046] FIG. 7 is a schematic diagram of the packet of FIG. 6A or 6B after being processed by the penultimate core switching node 50. It is assumed, for purposes of this example, that the egress port of the penultimate core switching node 50 connecting to the receiving egress edge switching node 52 is an untagged link.

[0047] The packet preferably includes an MPLs header data 128 including a destination address 120 that corresponds to the address of the egress edge switching node 52, a source address 122 that corresponds to the source address of the penultimate core switching node 50, and a protocol type 124 indicating that the packet is an MPLS packet. The MPLs header data 128 further includes the reserved virtual bridged LAN data unit label for indicating to the receiving egress edge switching node 52 that VLAN information is embedded within. Preferably the entire packet illustrated in FIG. 5 that is received from the EQM 62 is included as the MPLS payload data 129.

[0048] FIG. 8 is a flow diagram of a process for transmitting a data packet from the source host 40 to the target host 56 according to one embodiment of the invention. The process starts, and in step 130, the source host 40 transmits the packet to the ingress edge switching node 44. In step 132, the ingress edge switching node 44 processes the packet by attaching to it appropriate MPLS and VLAN classification information. The VLAN classification information is preferably attached even if the egress port used to forward the packet to the first hop is an untagged port.

[0049] In step 134, the processed packet is transmitted to a next hop on the LSP 58. In step 136, a determination is made if the penultimate core switching node 50 has been reached. If the answer is NC, the packet continues to be transmitted to the next hop, switching MPLS labels in a conventional manner, until the penultimate core switching node is reached.

[0050] In step 138, the penultimate core switching node, when reached, adds the reserved virtual bridged LAN data unit label to the received packet. In step 140, the processed packet is transmitted to the egress edge switching node 52. In step 142, the egress edge switching node retrieves the embedded original VLAN classification information and original source and destination address of the source host 40 and target host 56, respectively. In step 144, Layer 2 processing continues based on the retrieved information for forwarding the packet to the final destination, target host 56.

[0051] Although this invention has been described in certain specific embodiments, those skilled in the art will have no difficulty devising variations which in no way depart from the scope and spirit of the present invention. It is therefore to be understood that this invention may be practiced otherwise than is specifically described. Thus, the present embodiments of the invention should be considered in all respects as illustrative and not restrictive, the scope of the invention to be indicated by the appended claims and their equivalents rather than the foregoing description.

Claims

1. A virtual local area network (VLAN) tunneling system comprising: a first switching node having an ingress port and an egress port, the ingress port being associated with a VLAN and receiving a data packet, the egress port being configured as an untagged port, the egress port adding VLAN information to the data packet and transmitting the data packet on a label-switched path; a second switching node receiving the data packet transmitted from the first switching node via the label-switched path, the second switching node further identifying the data unit as a virtual bridged LAN data unit, retrieving the added VLAN information from the data packet, and transmitting the data packet to a final destination based on the retrieved VLAN information.

2. The VLAN tunneling system of claim 1 further including a third switching node in the label-switched path, the third switching node configured to add to the packet a label value reserved for packets originating from a port associated with a VLAN for informing the second switching node that VLAN information is embedded in the data packet.

3. The VLAN tunneling system of claim 2, wherein the second switching node receives the packet with the reserved label value and determines that the packet includes VLAN information associated with an originating port based on the reserved label.

4. A virtual local area network (VLAN) tunneling system including a switching node in a label-switched path, the switching node including: an ingress port associated with a VLAN receiving a packet; and an egress port configured as an untagged port, the egress port receiving the packet from the ingress port and adding to the packet VLAN information associated with the VLAN of the ingress port and transmitting the packet over the label-switched path.

5. A virtual local area network (VLAN) tunneling system including a switching node in a label-switched path, the switching node including: an ingress port receiving a packet having an ingress label value; and an egress port receiving the packet from the ingress port and replacing the ingress label value with a label value reserved for packets originating from a port associated with a VLAN, the egress port further transmitting the packet to a next switching node on the label-switched path based on the ingress label value.

6. The switching node of claim 5, wherein the next switching node is an egress edge switching node in the label-switched path.

7. The switching node of claim 6, wherein the egress edge switching node receives the packet with the reserved label value, determines that the packet includes VLAN information associated with an originating port based on the reserved label value, and processes the VLAN information for transmitting the packet to a final destination.

8. A virtual local area network (VLAN) tunneling system comprising: a first switching node having a plurality of ports at least one of which has a VLAN associated therewith; and a second switching node, characterized in that a data unit for transmission from said first switching node to said second switching node over said at least one port is checked for VLAN assignment prior to transmission and in that said data unit is transmitted from said first switching node to said second switching node on a label-switched path or not depending on a result of said check.

9. The system of claim 8 further characterized in that a VLAN identifier assigned to said data unit is applied to said data unit prior to transmission on said label-switched path.

10. The system of claim 9 further characterized in that said VLAN identifier applied to said data unit is referenced at said second switching node to identify the VLAN assignment of said data unit.

11. The system of claim 8, wherein said at least one port is untagged.

12. The system of claim 8 further characterized in that said data unit is bridged from said first switching node if not transmitted on said label-switched path.

13. A virtual local area network (VLAN) tunneling method comprising the steps of: receiving a packet at an ingress port, the ingress port being associated with a VLAN; forwarding the packet to an egress port, the egress port being configured as an untagged port; adding to the packet at the egress port VLAN information and a label value associated with a next switching node in a label-switched path; and transmitting the packet to the next switching node in the label-switched path.

14. A virtual local area network (VLAN) tunneling method comprising the steps of: receiving a packet at an ingress port of a switching node, the packet having an ingress label value; forwarding the packet to an egress port; replacing the ingress label value with a label value reserved for packets originating from a port associated with a VLAN; and transmitting the packet to a next switching node on a label-switched path based on the ingress label value.

15. The method of claim 16, wherein the next switching node is an egress edge switching node in the label-switched path.

16. The method of claim 15 further comprising the steps of: receiving at the egress edge switching node the packet with the reserved label value; determining that the packet includes VLAN information associated with an originating port based on the reserved label value; retrieving the VLAN information; and processing the VLAN information for transmitting the packet to a final destination.