JP2012085208A - Communication method of lisp network - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve routing or NAT traversal without concentrating the traffic even if the RLOC of xTR is a private address in an LISP network.SOLUTION: A registration procedure comprises: a procedure (t2) where the NAT encapsulates the registration request of transmission source information [EID1,RLOC1], describes the transmission source information in the header and transfers the encapsulated registration request to a mapping server; and a procedure (t3) where the mapping server MS/MR stores the transmission source information of the registration request as the RLOC(RLOC1) with the port number of EID1. The reception procedure has a procedure (t11) where a reception side xTR requests the mapping server for solution of destination address described in a response packet by EID base (t10), and a procedure (t12) where the response packet is transferred while encapsulating the destination address in the header of the RLOC(RLOC1) with the port number.

Description

  The present invention relates to a communication method of a LISP (Locator / ID Separation Protocol) network, and in particular, routing without causing traffic concentration even if an RTR (Routing Locator) of xTR (Tunnel Router) is a private address. And communication methods that can achieve NAT traversal.

  LISP (Locator / ID Separation Protocol) that realizes multi-home without any change to the host by performing packet encapsulation / decapsulation between the gateway routers (tunnel routers) at the site where the communicating host is located ) Is disclosed in Non-Patent Documents 1, 2, 3, and 4.

  FIG. 5 is a block diagram showing a typical configuration of the LISP network. The LISP site LNWs to which the transmission side host Hs is connected and the LISP site LNWd to which the reception side host Hd is connected are respectively tunnel routers ITR. It is connected to the Internet of the core network via (Ingress Tunnel Router) and ETR (Egress Tunnel Router). The sending host Hs and the receiving host Hd are assigned EID1 and EID2 as host identifiers (EID: Endpoint Indentifier), respectively, and each of the tunnel routers ITR and ETR is assigned RLOC1 as a location identifier (RLOC: Routing Locator). , RLOC2 is assigned.

  A mapping server MS for mapping the host identifiers EID1, EID2 and the location identifiers RLOC1, RLOC2 is installed on the Internet. The mapping server MS has a simple xTR (generic name for ITR and ETR) function, and the IP address used as RLOC is often a global address.

  In such a LISP network, routing is performed based on the host identifier EID in each LISP site, and routing is performed based on the EID location identifier RLOC in the core network. In the ITR that exists at the boundary between the LISP site LNWs on the sending side and the core network, the header used for routing in the core network is added, and in the ETR that exists at the boundary between the LISP site LNWd on the receiving side and the core network, The header is deleted.

  In such a LISP network, when a private address is used as RLOC, there is a problem that the receiving side cannot reply to the transmitting side because routing and NAT traversal cannot be performed. FIG. 6 is a diagram for explaining the technical problem of such a LISP network.

  When the sending host Hs generates an IP packet and sends it to the receiving host Hd, the sending host Hs has a simple xTR function, so the address information is EID-based [Src: EID1, Dst: EID2]. A pair of RLOC and port number [Src: RLOC1, Sport: 10000, Dst: RLOC2, Dport: 4341] is added to the transmission packet as an Outer header for LISP and encapsulated. The port number [4341] is a standard port number assigned in advance for the LISP network.

  The NAT function on the transmission side rewrites the transmission source address and port number of the Outer header into its own global address and port number pair [Src: IP1, Sport: 5678] and transfers the result to the Internet.

  The receiving host Hd decapsulates the received packet, extracts the transmission packet, generates a response packet in which the address information is described in EID base [Src: EID2, Dst: EID1], and returns the response packet. xTR (RLOC2) encapsulates the response packet with an LISP Outer header [Src: RLOC2, Sport: 12000, Dst: RLOC1, Dport: 4341] and transfers it.

  However, since the destination address RLOC1 is a private address, it cannot be routed and communication is interrupted. Even if this packet reaches the NAT function, the Dport is different from the output port number [Sport: 5678] of the NAT, and is not regarded as the same session and is discarded by the NAT function.

  Non-patent document 5 discloses an NTR (NAT Traversal Router) that solves such technical problems related to NAT traversal and can realize NAT traversal even if the location identifier RLOC1 is a private address in a LISP network. This NTR achieves NAT traversal by the following procedure.

  Procedure 1: Instead of the IP address (private) of xTR behind NAT, the IP address (global) of NTR is set as RLOC1, and this is registered with the mapping server MS in association with the EID of xTR. As a result, a packet addressed to xTR is guided to NTR.

  Procedure 2: xTR behind NAT performs Map-Register for NTR using source port number [4341]. The NTR records a pair of a source IP address and a port number after NAT conversion.

  Procedure 3: When the NTR receives a packet addressed to the xTR behind the NAT, it rewrites the destination IP address and the destination port number to the contents recorded in the procedure 2 and transfers them.

LISP, http://tools.ietf.org/html/draft-ietf-lisp-07 LISP-MS, http://tools.ietf.org/html/draft-ietf-lisp-ms-05 LISP-MN, http://tools.ietf.org/html/draft-meyer-lisp-mn-01 LISP-MN-NAT, http://tools.ietf.org/html/draft-klein-lisp-mn-nat-traversal-00 Dominik Klein, Matthias Hartmann, and Michael Menth, NAT Traversal for LISP Mobile Node, 2010

  In the non-patent document 5, the first technical problem that the response packet cannot be routed in the core network if the RLOC is a private address is solved by substituting the address of the NTR device with the RLOC1. In addition, in the LISP network, since the destination port number of the data packet is always [4341], the first technical problem (NAT exceeded) is that the destination port number of the packet is changed in the NTR device. It has been resolved.

  However, in the above prior art, all traffic destined for xTR behind NAT is directed to NTR, so traffic is concentrated on NTR, and NTR needs to rewrite the destination address and port number of all IP packets. Therefore, the processing load becomes heavy. Therefore, a large number of NTRs must be installed in order to distribute the load, and if the NTRs are installed at geographically distant locations, the delay will increase.

  An object of the present invention is to solve the above-described problems of the prior art and provide a communication method capable of realizing routing and NAT traversal without causing concentration of traffic even when RLOC is a private address in a LISP network. .

  In order to achieve the above object, according to the present invention, the LISP sites on the transmission side and the reception side are connected to the core network including the mapping router via the tunnel router xTR, and the host identifier EID is included in each LISP site. The LISP network communication method in which routing is performed on the basis of the core network and routing is performed on the basis of the EID location identifier RLOC is characterized by the following procedure.

  (1) Registration procedure:

  In the sending side xTR, a procedure for encapsulating a registration request in which the correspondence between EID and RLOC of the sending side host is described in an Outer header in which address information is described in RLOC base, and in the NAT on the sending side, the Outer header Rewriting the source address of the IP address to a pair of output port address and output port number of NAT and transferring it to the mapping server, and storing the correspondence relationship between the addresses before and after the rewriting in a NAT table, and the mapping server sends the registration request And storing the NAT output port address and port number pair as RLOC with a port number in association with the EID of the sending host.

  (2) Transmission procedure:

  A procedure for generating a transmission packet in which address information is described in EID base in the transmission side host; and a procedure in the transmission side xTR for encapsulating the transmission packet in an Outer header in which address information is described in RLOC base; The NAT includes a procedure of rewriting and transmitting a source address described in the RLOC base in the Outer header to a pair of the NAT source address and source port number.

  (3) Reception procedure:

  The receiving side xTR receives and decapsulates the encapsulated transmission packet, and the receiving side host, based on the decapsulated transmission packet, generates a response packet in which address information is described in EID base. A procedure for generating, a procedure for requesting the mapping server to resolve a destination address described in EID base in the response packet in the receiving side xTR, and in response to the address resolution request in the mapping server, Procedure for responding to RLOC with port number associated with destination address described in EID base, and forwarding packet encapsulated in Outer header of destination RLOC with port number at destination xTR And receiving the encapsulated response packet in the NAT and receiving the destination There are a procedure for rewriting and transferring the address based on the RLOC based on the NAT table, and a procedure for decapsulating and transferring the encapsulated response packet at the transmission side xTR.

  According to the present invention, routing and NAT traversal can be realized without concentrating traffic on some nodes even if the RLOC address of xTR is a private address in a LISP network.

1 is a block diagram showing a configuration of a LISP network to which the present invention is applied. It is a functional block diagram for demonstrating the communication method of the LISP network which concerns on one Embodiment of this invention. 5 is a sequence flow for explaining a communication method of a LISP network according to an embodiment of the present invention. It is the block diagram which showed the structure of other embodiment of this invention. It is the block diagram which showed the structure of the conventional LISP network. It is a block diagram for demonstrating the subject of the conventional LISP network.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a main part of a LISP network to which the present invention is applied. On the Internet as a core network, a mapping server MS / MR (Map Server / Map Resolver) for address resolution is shown. The xTR function on the transmission side is implemented in the transmission side host Hs on the back side of the NAT device as viewed from the core network.

  The NAT is a full cone NAT, which holds only the translation pair of the NAT device's back side (inside) address (iAddr: iPort) and the front side (outside) address (eAddr: ePort). Since the packet from the host to eAddr: ePort is forwarded to iAddr: iPort, the open port can be accessed from any node. Such full cone NAT is described in detail in STUN (http://tools.ietf.org/html/rfc3489).

FIG. 2 is a functional block diagram for explaining a communication method of a LISP network according to an embodiment of the present invention, and FIG. 3 is a sequence flow thereof. Here, the location identifier RLOC1 of xTR mounted on the sending host Hs is a private address, the location identifier RLOC2 of xTR mounted on the receiving host Hd is a global address, and the mapping server MS / MS has a global address Explanation will be given assuming that IP2 is assigned.

1. Registration Procedure Referring to the sequence flow of FIG. 3, at time t1, a mapping registration (Map-Register) request message in which xTR (RLOC1) implemented in the sending host Hs is embedded with source information [EID1, RLOC1] Is encapsulated with an Outer header in which address information is described in RLOC base and transmitted to the mapping server MS / MR. In the Outer header, [Src: RLOC1, Sport: 4341] is described as a pair of source address and source port number, and [Dst: IP2, Dport: 4342] is described as a pair of destination address and destination port number Has been.

  At time t2, the NAT device sets the source address and source port number [Src: RLOC1, Sport: 4341] of the Outer header of the mapping registration request message as a pair of the output port address and output port number of the NAT device [ Src: IP1, Sport: 2345] is rewritten and transferred to the mapping server MS / MR, and the correspondence between the address and port number before and after the rewrite is stored in the NAT table.

  At time t3, the mapping server MS / MR that has received this mapping registration request message detects that the already registered mapping information does not match the newly received mapping information with respect to RLOC1, and the request is related to NAT. Judged as a mapping request. Then, the source address and source port number pair [Src: IP1, Sport: 2345] is stored as RLOC (RLOC1) with a port number of EID1.

At time t4, the receiving side xTR transmits a mapping registration request message [EID2, RLOC2] to the mapping server MS / MR, and this is registered in the mapping server MS / MR as in the prior art.

2. Transmission Procedure Referring also to FIG. 2, at time t5, the transmission side host (EID1) application converts the source and destination address information to EID base [Src: EID1, Sport: 1000, Dst: EID2, Dport: 80. ] Is generated. At time t6, the transmission packet is transmitted to the source address Src and the source port number Sport [RLOC1, 10000], the destination address Dst and the destination port number Dport are set to “RLOC2, 4341 "is encapsulated in the Outer header and transferred to the NAT device.

At time t7, in the NAT device, the source address and source port number pair “RLOC1,10000” in the Outer header is rewritten to the output port address and output port number pair [IP1,5678] of the NAT device. Forwarded. At time t8, the packet is received and decapsulated by the receiving side xTR (RLOC2), and the transmission packet [Src: EID1, Sport: 1000, Dst: EID2, Dport: 80] is extracted and received by the receiving side host Hd. Transferred to application (EID2).

3. Reception procedure

  At time t9, the application on the receiving host Hd returns a response packet in which the address information of the source and destination is described in EID base [Src: EID2, Sport: 80, Dst: EID1, Dport: 1000] and xTR (RLOC2). At time t10, the xTR (RLOC2) requests the mapping server MS / MR to resolve the address of the destination EID1. At time t11, the mapping server MS / MR refers to the mapping information and responds with RLOC [IP1, 2345] with a port number associated with the EID1.

  At time t12, xTR (RLOC2) sends the response packet [Src: EID2, Sport: 80, Dst: EID1, Dport: 1000], a source address and source port number pair [RLOC2, 12000], and a destination address And a pair of destination port numbers are encapsulated in an Outer header with the port number RLOC [IP1, 2345] and transferred.

  At time t13, the response packet is received by the transmission-side NAT device (IP1) in which the port of port number [2345] is open, and the destination address and destination port number pair “IP1, 2345” in the Outer header is set. Based on the NAT table, the address is rewritten to the RLOC base address [Src: RLOC1, Sport: 4341].

  At time t14, the received packet is decapsulated and the response packet [EID2, S0, EID1, 1000] is extracted and transferred to the application (EID1) by the xTR function implemented in the transmission side host Hs.

  In the above embodiment, the case where the transmission side host is connected to the back side of the NAT as viewed from the core network on the transmission side has been described as an example, but the present invention is not limited to this, As shown in FIG. 4, the present invention can be similarly applied to a case where a LISP site is formed on the back side of NAT and a transmission side host is connected to the LISP site.

  Hs: Sending host, Hd: Receiving host, xTR: Tunnel router

Claims (4)

Each LISP site on the transmission side and the reception side is connected to the core network equipped with the mapping router via the tunnel router xTR, routing is performed based on the host identifier EID in each LISP site, and EID of the core network is performed. In a communication method of a LISP network where routing is performed based on the location identifier RLOC,
The registration procedure
In the sending side xTR, a procedure for encapsulating a registration request in which the correspondence between EID and RLOC of the sending side host is described in an Outer header in which address information is described in RLOC base, and
In the NAT on the transmission side, the source address described in the RLOC base in the Outer header is rewritten to a pair of NAT output port address and output port number and transferred to the mapping server, and the correspondence relationship between the addresses before and after the rewrite is shown. The procedure to store in the NAT table,
In the mapping server, the registration request is received, and a pair of the NAT output port address and port number is stored as an RLOC with a port number in association with the EID of the sending host, and
Send procedure is
In the sending host, a procedure for generating a send packet in which address information is described in EID base, and
In the transmission side xTR, a procedure for encapsulating the transmission packet with an Outer header whose address information is described in RLOC base, and
In the NAT, the source address described in the RLOC base in the Outer header has a procedure of rewriting and transferring the source address and source port number pair of the NAT,
The receiving procedure is
In the receiving side xTR, a procedure for receiving and decapsulating the encapsulated transmission packet;
In the receiving host, based on the decapsulated transmission packet, a procedure for generating a response packet in which address information is described on an EID basis;
In the reception side xTR, a procedure for requesting the mapping server to resolve the destination address described in EID base in the response packet;
In the mapping server, in response to the address resolution request, a procedure of responding with an RLOC with a port number associated with the destination address described in the EID base;
In the receiving side xTR, the response packet is encapsulated in the Outer header of the RLOC with the port number as the destination number and transferred,
In the NAT, receiving the encapsulated response packet, rewriting and transferring the destination address information to the RLOC base based on the NAT table, and
A communication method for a LISP network, comprising: a step of decapsulating and transferring the encapsulated response packet in the transmission side xTR.   The LISP network communication method according to claim 1, wherein a transmission side host is connected to the back side of the NAT on the transmission side as viewed from the core network.   The LISP network communication method according to claim 1, wherein a LISP site is formed on the back side of the NAT as viewed from the core network on the transmission side, and a transmission side host is connected to the LISP site.   4. The LISP network communication method according to claim 1, wherein the NAT is a full cone NAT.

Images