WO2003041365A1

WO2003041365A1 - Dynamic allocation of a temporary ipv4 address to an ipv6 equipment

Info

Publication number: WO2003041365A1
Application number: PCT/FR2002/003706
Authority: WO
Inventors: Gilles Diribarne; Philippe Bereski
Original assignee: Alcatel
Priority date: 2001-11-05
Filing date: 2002-10-29
Publication date: 2003-05-15
Also published as: FR2832012A1

Abstract

The invention relates to the dynamic allocation of a temporary IPV4 address to an IPV6 equipment. More specifically, the invention relates to a name server (DNS) which comprises means of receiving name requests from a transmitting piece of equipment and which contains the name of a receiving piece of equipment. A first equipment (Ev4) among said transmitting and receiving pieces of equipment conforms solely to a first protocol, while the second equipment (Ev6) conforms to said first protocol and to a second protocol. The inventive server is characterised in that it also comprises means of detecting if the address of the second equipment is not compatible with the first equipment and, in this case, of triggering the dynamic allocation of an address which is compatible with the first equipment to said second equipment.

Description

Dynamic assignment of a temporary IPv4 address to an IPv6 device

The present invention relates to the transition mechanisms from version 4 of the internet protocol (IPv4) to version 6 of this protocol (IPvό).

More particularly, it relates to the transient situation where data networks conforming to the two protocols coexist and must communicate.

Existing IPv4 devices have no simple way of being brought into compliance with the new IPvό protocols. It is therefore necessary to define mechanisms allowing these IPv4 devices to communicate with new IPvό devices.

It is generally accepted that, at least during this transitional period, equipment conforming to the IPvό protocol will also comply with the IPv4 protocol, in order to maintain backward compatibility. These devices will therefore have two protocol stacks, an IPv4 and an IPvό. This type of equipment is generally known by the English term "dual stacked".

However, due to the scarcity of IPv4 addresses, these dual stacked devices will be forced to share, over time, the same subset of IPv4 addresses.

The problem then arises when an IPv4 device wants to initiate communication with an IPvό / IPv4 device (that is to say having the two protocol stacks) to which no IPv4 address has been assigned, or vice versa. IPv4 equipment, in fact, can only include IPv4 addresses and, in general, the network to which it belongs, can only carry data packets conforming to the IPv4 protocol. A solution was provided by the DSTM mechanism (Dual Stock

Transition Mechanism) which consists of dynamically assigning an address

IPv4 to IPv6 / IPv4 equipment, by means of a dynamic DHCP address server managing a set (or pool, according to terminology in English) of available IPv4 addresses.

Within the Internet Engineering Task Force (IETF), this mechanism is in the draff document stage, under the names "draft-ietf-ngtrans-dstm.txt" and "draft-ietf-ngtrans-aaic.txt", in one of their successive versions.

The DSTM mechanism as presented in this document is illustrated in Figure 1.

In the case of E _v4 equipment of an N _v4 network conforming to IPv4 wishing to communicate with an E _v6 equipment having a double IPv4 / IPv IP protocol stack and belonging to an N _v6 network conforming to IPvό, the assignment of a temporary IPv4 address to the E _{v6 device} , takes place according to the following steps:

• The E _{v4 device} transmits a name request to a name server (or DNS for Domain Name Server, in English) associated with the N _v6 network to which the E _v6 device belongs.

• Noting that node E _v6 does not have an IPv4 compatible address, the DNS name server then transmits a request to a server dedicated to dynamic address allocation, or DHCP (Dynamic Host Configuration Protocol) server, which is associated with this network N _v6 .

• The DHCP server transmits to the E _{v6 device} , a message indicating that it must reconfigure itself.

• The E _u6 equipment responds with a reconfiguration request, transmitted to the DHCP server. • The DHCP server then transmits a temporary address to the E _{v6 device} . At the same time, it transmits update messages on the one hand to the DNS name server, and on the other hand to the gateway G which connects the networks N _v4 and N _v6 . • Finally, the name server responds to the E _{v4 device} , with the temporary address which has been assigned to the E _{v6 device} .

This E _v4 equipment can then communicate with the E _v6 equipment.

Such a method has several drawbacks: It requires, in particular, a large number (8, in fact) of messages to be transmitted between the various pieces of equipment involved in the mechanism. These messages can contribute to network congestion, or at least to a loss of performance due to excessive load. In addition, in its current implementation, this process requires the complex services of a dynamic DHCP configuration server subject to numerous criticisms within the Internet standardization body (IETF). In reality, the use of DHCP comes down to dynamic address management. This function alone has no reason to be in an IPvό network, because the protocol integrates it intrinsically.

The state-of-the-art solution therefore requires the use of additional equipment which could be considered useless outside of this use. The addition of this equipment being expensive, such a solution is to be avoided. An extension of this mechanism has also been proposed, under the name

"Draft-ietf-ngtrans-dstmextl-aiih-OO.txt". However, this extension is also based on the use of a dynamic configuration server and therefore suffers from the same drawbacks as those mentioned above. The object of the invention is to propose a mechanism allowing an IPv4 device to initiate a communication with an IPv6 / IPv4 device having no fixed IPv4 address, and vice versa, which does not require a dynamic address server. DHCP type.

To do this, the invention relates to a name server having means for receiving requests for names from a sending equipment and containing the name of a receiving equipment, a first equipment among said sending and receiving equipment, being conforms only to a first protocol, the second equipment being conforms to said first protocol and to a second protocol. This name server is characterized in that it also has means for detecting that the second device does not have an address compatible with the first device and, in this case, triggering the dynamic assignment, to the second device, of '' an address compatible with the first device.

According to an implementation of the invention, the dynamic assignment is carried out directly by the name server according to a set of available addresses compatible with the first device. The name server then further comprises means for notifying the address dynamically assigned to the second device and / or to the gateway (G) allowing communication between the first device and the second device and / or to the sending device. in case it is different from the second equipment.

According to another implementation of the invention, the dynamic assignment is carried out by sending a request for address assignment to the gateway allowing communication between the first device and the second device. The invention also relates to such a gateway, that is to say a gateway allowing communication between a first network conforming to a first protocol and a second network conforming to a second protocol. This gateway is characterized in that it has a set of available addresses and in that it has means for receiving requests for address allocation from a name server, and transmitting an address chosen from this set, to the name server and to the equipment whose address is contained in the request for address assignment.

According to a preferred application of the invention, the first device complies with the IPv4 protocols and the second device complies with the IPvό protocols and the IPv4 protocol.

In addition to the advantage of being able to save on a dynamic configuration server of DHCP type, the invention also has the advantage of significantly reducing the number of messages exchanged before arriving at the address assignment.

The invention and its advantages will appear more clearly in the description of implementations which will follow, in conjunction with the appended figures.

FIG. 1, already commented on, represents the DSTM mechanism of the state of the art.

FIG. 2 illustrates the communication on the initiative of the IPv4 equipment according to a first implementation of the invention.

Figures 3a and 3b show the communication on the initiative of the dual protocol stack IPv4 / IPv6 equipment, according to this same implementation.

FIG. 4 illustrates the communication on the initiative of the IPv4 equipment according to a second implementation of the invention. Finally, Figure 5 presents the communication on the initiative of the dual protocol stack IPv4 / IPv6 equipment, according to this second implementation.

In Figure 2 are shown two networks N _v4 and N _v6 connected by a gateway. The N _v4 network complies with the IPv4 protocol while the N _v6 network complies with the IPvό protocol. The gateway G is in charge of carrying out packet encapsulations and decapsulations, in order to allow the transmission of packets between these two networks.

The network N _v4 includes an equipment E _v4 which wants to initiate a communication with an equipment E _v6 of the network N _v6 . This E _v6 equipment has a double stack of IPv4 and IPvό protocols. It obviously has an IPvό address, but no IPv4 address.

Firstly, the equipment E _v4 transmits a request for names m, to the DNS name server of the network N _v6 . This name query is classic and conforms to the usual mechanism. It conveys the name of the E _{v6 device} whose E _{v4 device} wants the address. It is also known that this name request can pass through a functional layer called “resolver” in English, which is local to the N _v4 network and which has the function of determining the appropriate DNS name server and of transmitting the request for last name.

The DNS name server has means for detecting that this E _v6 equipment does not have an address compatible with the E _v4 equipment, that is to say an IPv4 address.

If this is the case, it triggers the dynamic assignment of a temporary (or dynamic) IPv4 address to this IPvό device. According to the implementation illustrated in FIG. 2, this dynamic assignment takes place by transmitting a request for assignment of address m ₂ to the gateway G.

This request for address assignment m ₂ contains the IPvό address of the equipment E _v6 .

According to an implementation of the invention, this message m ₂ can be a message conforming to the ICMP protocol (Internet Control Message Protocol).

This gateway G has a set of available IPv4 addresses.

Upon receipt of this request for address assignment, it can therefore choose an address available from this set.

This choice is memorized by the gateway G, on the one hand so as not to reassign this address during another name request, and on the other hand to be able to correctly transmit the data streams that the equipment E _v4 will send to the E _v6 equipment.

In response to the request for assignment of address m ₂ , the gateway G sends two messages: A first message m ₃ is transmitted to the equipment E _v6 , whose address IPvό has been transmitted in the request message m ₂ in address assignment. This message m ₃ contains the temporary address which has been assigned to it.

A second message m ₄ is a response message to the name request, which is transmitted to the DNS name server. This message m ₄ also contains the temporary address.

This response message m ₄ can, according to an implementation of the invention, be a message conforming to the ICMP protocol (Internet Control Message Protocol). The DNS name server can remember this temporary address assignment. So if a new name request is made to it

(by another device of the N _v4 network, for example), this temporary address can be reused without a new dynamic allocation mechanism being triggered.

The DNS name server can insert the temporary address in the response message which it then transmits to the E _{v4 device} .

The E _{v4 device} now having a temporary IPv4 address of the E _{v6 device} can transmit a data stream to it. This data flow passes through the gateway G. This gateway G having stored this temporary address IPv4 can correctly route it in the network N _v6 so that it actually arrives at the equipment E _vό . This gateway can use, to do this, the encapsulation and de-encapsulation mechanisms from one protocol to another. This mechanism can be in accordance with that described in RFC 1241 entitled "Scheme for an Internet encapsυlation protocol".

Figures 3a and 3b show the same equipment as Figure 2, but to illustrate the mechanism implemented when it is the equipment E _v6 which wants to initiate communication with the node E _v4 . Two possibilities can be implemented:

The first possibility is illustrated in Figure 3a. According to this implementation, when the DNS name server receives the request for name m, transmitted by the equipment E _v6 , it responds to it with a message m ₂ '. On receipt of this message m ₂ ′, the equipment E _v6 will send a message m ₃ ′ to the gateway G and trigger the same mechanism as previously: assignment of a dynamic address by the gateway G then transmission of the messages m ₃ and m ₄ to the DNS name server and to the E _v6 equipment. The second possibility, illustrated in Figure 3b, is to give proxy power to the DNS name server. On receipt of the name request m ,, the DNS name server directly notifies the gateway G, by a message m, ", which can then trigger the mechanism described above for dynamic address assignment and notification to the server. DNS names and E _v6 equipment, by two messages m ₃ and m ₄ .

FIG. 4 illustrates another implementation of the invention, in which the dynamic address assignment function is supported by the DNS name server.

When the equipment E _v4 of the network N _v4 wishes to initiate a communication with the equipment E _v6 of the network N _v6 , it first transmits a request for name m, to the name server DNS. As before, this name request can be relayed by an application layer of the network N _v4 usually called "resolver" and in charge of transmitting the name request to the DNS name server which belongs to the network N _v6 .

This name request, in accordance with the usual mechanisms and protocols, conveys the name of the E _{v6 device} . The DNS name server has means for detecting that this equipment does not have an address compatible with the E _v4 equipment, that is to say an IPv4 address. In this case, it can trigger the dynamic assignment of a dynamic IPv4 address to this IPvό device.

To do this, the DNS name server has a set of available IPv4 addresses. Upon receipt of this name request, he can therefore choose an available address from this set. This choice can then be stored in the name / address correspondence tables of the name server.

It can then transmit this dynamic address: • On the one hand to the gateway G, in the form of a message m ₂ , • On the other hand to the E _v6 equipment in the form of an m ₃ message.

As it was said previously for the first implementation, these messages can for example comply with the Internet Control Message Protocol (ICMP).

The case where it is the equipment E _v6 which wishes to initiate the communication with the equipment E _v4 is illustrated by FIG. 5. It begins with the request for name m, transmitted by the equipment E _v6 to the name server DNS . The latter noting that the node E _v6 does not have an IPv4 compatible address, implements the dynamic allocation mechanism described above. It then notifies this dynamic address:

• On the one hand to the gateway G, in the form of a message m ₂ ,

• On the other hand to the E _v6 equipment in the form of an m ₃ message.

Claims

1) Name server (DNS) having means for receiving name requests from a sending equipment and containing the name of a receiving equipment, a first equipment (E _v4 ) among said sending and receiving equipment, being compliant only to a first protocol, the second equipment (E _v6 ) conforming to said first protocol and to a second protocol, characterized in that it also has means for detecting that said second equipment does not have an address compatible with said first equipment and, in this case, trigger the dynamic assignment, to said second equipment, of an address compatible with said first equipment.

2) Name server according to claim 1, in which the dynamic assignment is carried out directly by said name server according to a set of available addresses compatible with said first item of equipment, and further comprising means for notifying the address dynamically assigned to said second device and / or to the gateway (G) allowing communication between said first device and said second device and / or to the sending device in the case where the latter is different from said second device.

3) Name server according to claim 1, in which the dynamic assignment is carried out by sending a request for address assignment to the gateway allowing the communication between said first device and said second device.

4) Name server according to one of claims 1 to 3, wherein said first device complies with IPv4 protocols and said second device complies with IPvό protocols and IPv4 protocol. 5) Gateway (G) allowing communication between a first network (N _v4 ) conforming to a first protocol and a second network (N _vό ) conforming to a second protocol, characterized in that it has a set of addresses available and in that it has means for receiving requests for address allocation from a name server (DNS), and transmitting an address chosen from said set, to said name server and to an equipment including address is contained in said request for address assignment.