From MBone to M6Bone

Transcription

1 From MBone to M6Bone Assistant Lecturer Madalina Mlak Academy of Economic Studies Bucharest, Faculty of Economic Cybernetics, Statistics, and Informatics Abstract This paper present a parallel between networks MBone and M6Bone, the connection between those networks and hosts, the map of sites connected to M6Bone network. Keywords: MBone, M6Bone, IPv6, multicast, routing protocols. 1. Parallel MBone-M6Bone 1.1. MBone origin In 1992 was born MBone (Multicast Backbone). The MBone created from experiments during IETF (Internet Engineering Task Force) meeting in which live audio and video were transmitting around the world What is the MBone? The MBone network is a group of sites (a network of hosts) connected to the Internet which communicating using a technique called IP multicast and using to develop protocols and applications. MBone allows multicast packets to travel through routers that are set up to handle only unicast traffic. In order to participate in the MBone network it is necessary to have a workstation supporting IP multicast and to have a network connection with a reasonable bandwidth (typically around 1 Mbps). The necessary of IP multicast and software applications in freely is available How large is the MBone? The size of MBone compared to the Internet as a whole is relatively small (~ a few percent of the Internet) MBone topology Within a continent, the MBone topology will be a combination of mesh and star: the backbone and regional network will be link by mesh of tunnels. Some redundant tunnels may be configuring with higher metrics for robustness. Then each regional network will have a star hierarchy hanging off are node of the mesh to fan out and connect too all the customer networks that want to participate. Between continents, there will probably be only one or two tunnels, preferably terminating at the closet point on the MBone mesh M6Bone origin This project which started in July 2001 and the scope is to offer an IPv6 multicast service for sites. This service is based on Renater3 (IPv6 enable network), on G6 group (French group of IPv6 testers) and benefits from the logistic support of the Aristote Association which is involved in the broadcasting of the modern technologies What is the M6Bone? M6Bone is an experimental IPv6 Multicast network. M6Bone allow people who are connect to learn about IPv6 multicast and M6Bone is a good place to test IPv6 multicast equipments, configurations, implementations or software and to develop new protocols and services M6Bone topology The M6Bone have to use a different topology for IPv6 multicast and IPv6 unicast traffic. If the IPv6 multicast and unicast topologies are not the same, the IPv6 multicast routing protocol performs the RPF check using the multicast routing table. In actual IPv6 native networks, multicast routing is not available on the routers. That is why this project 119

2 decided to develop the M6Bone with tunnel architecture with edge equipments supporting IPv6 multicast. Today, Renater provides a central router that is pleased to welcome sites willing to connect. The way to connect to the M6Bone is to create a tunnel between an existing M6Bone router (for example, Renater) and a site. for sites that already have an IPv6 connectivity, the tunnel will be an IPv6 (multicast) in IPv6 (unicast) tunnel; for sites that only have an IPv4 connectivity, the tunnel will be an IPv6 (multicast) in IPv4 tunnel. IPv6 in IPv6 and IPv6 in IPv4 tunnels are deploy to connect M6Bone routers How large is the M6Bone? Today over 35 sites in Europe and beyond are connect to the M6Bone, as shown in word map The world and European map The world and European map of the M6Bone are present in the following pictures: 120

3 1.10. Sites connected to the M6Bone Examples of sites connected to the M6Bone. List of the connected sites is much more but in the following are present some of them: France: GIP Renater, Paris Aristote, Paris DESS ART, Paris IRISA/INRIA, Rennes Université de Bretagne sud, Vannes Université Louis Pasteur, Strasbourg Belgium: Université Libre de Bruxelles Luxembourg: RESTENA Nederland: Surfnet Finland: Oulu University Oulu Polytehnic Norway: Osfold College Germany: JOIN Poland: Poznan Supercomputing and Networking Austria: Vienna University Computer Center Spain: UPM, Madrid University of Murcia Italy: UK: CNR-IIT University College of London Dante, Cambridge University of Southampton SUA: Philippines:ASTO, Quezon Taiwan: ASCC Senegal: ESMT, Dakar Korea: ETRI Mexico: UdeG, Guadalajara Wichita State University, Wichita, Kansas Collaboration for M6Bone Connect to the M6Bone and join the mailing list: Send an to with message "subscribe m6bone". 2. IP multicast 2.1. IP multicast addresses Every IP multicast group has a group address IP multicast provides only open groups. That is it is not necessary to be a member of a group in order to send datagrams to the group. Multicast address is like IP addresses used for single hosts and it is writing in the same way. Multicast addresses will never have conflict with host addresses because a part of the IP address space is specifically reserve for multicast. The multicast addresses are in the range from to However, the multicast addresses from to are reserve for multicast routing information. Applications programs should use multicast addresses outside this range. Multicast address assignment is generally dynamic and under the control of collections of the users Class D address A multicast IP packet has a class D address used for multicast (from to ). Class D (Table 1.) has a first bit value of one and second bit value of 1, third bit value of 1 and fourth bit value of 0. The other 28 bits are using to identify the group of computers the multicast message is intending for these. Class D accounts for 1/16 th (268,435,456 or 2 28 ) of the available IP addresses Net Host or Node Tabel 1. Class D address There is no restriction on the physical location or number of members in a multicast group and it possible that they may be members of more that one multicast group. IP multicast packets are encapsulating for sending through tunnels, so that they look like normal unicast datagrams to intervening routers and subnets. A multicast router, which wants to transmit a multicast packet across a tunnel, will prep ending another IP header. A multicast router set the destination address in the new header to be a unicast address of the multicast router at the other end of the tunnel and set the IP protocol field in the new header to be four, which means the next protocol is IP IPv6 addressing architecture IPv6 addresses are 128-bit identifiers for interfaces and sets of interfaces. There are three types of addresses: unicast: an identifier for a single interface. A packet sent to a unicast address is delivered to the interface identified by that address; anycast, multicast: identifiers for a set of interfaces (typically belonging to different nodes); broadcast: is replaced by multicast. Multicasting is a technical term, a network routing facility that means a method of sending packets to multiple sites at the same time. How big a packet is depends on the protocol involved. Multicasting is the ability of the network to efficiently delivering information to multiple recipients. 121

4 2.4. Unicast-prefix based IPv6 multicast address An extension to the multicast addressing architecture that allows for unicast-prefix-based allocation of multicast addresses. By using those types of addresses, network operators will be able to identify their multicast addresses without need to run an inter-domain allocation protocol. flags is a set of 0 0 P T: if P=0 indicates a multicast address that is not assigned based on the network prefix. This indicates a multicast address as defined in RFC2273; if P=1 indicates a multicast address that is assigned based on the network prefix; if P=1, T must be set to 1, otherwise the setting of the T bit is defined in section 2.7 of RFC2273 the reserved field must be zero. plen indicates the actual number of bits in the network prefix field that identify the subnet when P=1 network prefix identifies the network prefix of the unicast subnet owning the multicast address. If IP=1, this field contains the unicast network prefix assigned to the domain owning, or allocating, the multicast address. All non-significant bits of the network prefix field should be zero group ID is set based on the guidelins outlined in RFC Ipv6 multicast address Structure of IPv6 multicast address is presents in the following table: bits flags scope group ID Table 2. Structure of IPv6 multicast address =FF: at the start of the address identifies the address as being a multicast address flags (0 0 0 T): if T=0,indicates a permanentlyassigned ( well-known ), and must be initialized to 0; if T=1,indicates a non-permanently assigned ( transient ) multicast address scope of the multicast group and the value are: 0,F reserved, 1 node-local scope, 2 link-local scope, 3,4,6,7,9,A,D unassigned, 5 site-local scope, 8 organization-local scope, E global scope. group IP identifies the multicast group, either permanent or transient, within the given scope. IANA (Internet Assigned Numbers Authority) handles both fixed and variable scope for multicast address allocations IPv6 anycast address Structure of IPv6 anycast address is presents in the following table: flags scope reserved plen network group prefix ID Table 3. Structure of IPv6 anycast address an IPv6 anycast address is an address that is assigned to more than one interface (typically belonging to different nodes); a packet sent to an anycast address is routed to the nearest interface having that address, according to the routing protocols measure of distance; anycast addresses are allocated from the unicast address space, using any of the defined unicast address formats; for any assigned anycast address there is a longest address prefix P that identifies the topological region in which all interfaces belonging to that anycast address reside. 3. IPv6 multicast hosts An IPv6 multicast host has: to have an IPv6 stack that supports MLD (Multicast Listener Discovery); to be to run mbone tools. Many tools are already available and can be install on almost all operating systems having IPv6 stack. Examples of available operating systems, which support IPv6: FreeBSD Linux or Red Hat Linux Windows2000 or Windows XP For example to install mbone tools on Windows XP. The operating systems configuration follows those steps: 1. first install the Service Pack1 2. second install the IPv6 stack 3. after that open a DOS command prompt and execute the command: ipv6 install To install mbone tools on Windows XP download the software on (Microsoft part). To run mbone tools on Windows XP your computer must be in the IPv6 DNS of your site. 122

5 For Windows 2000 download and install the IPv6 stack from sdks/platform/tpipv6.asp. To run mbone tools on Windows 2000 your computer must be in the IPv6 DNS of your site. 4. IPv6 routing and tunnels The PIM (Protocol Independent Multicast) protocol uses unicast routing table for RPF (reversepath-forwarding) check. It implies that unicast and multicast topologies must be the same. It is not the case since the M6Bone is an overlay network. A solution to this problem is to use a multicast routing table but it is not yet implement on IPv6 multicast routing equipment. The sites have to use separate routers for unicast and multicast. The IPv6 multicast tunnels are setup between IPv6 multicast routers. Those routers exchange their unicast routing table that will be use for RPF check using RIPng (Routing Information Protocol next generation) protocol. With RIPng, each site will advertise its prefix corresponding to the subnet where IPv6 multicast is enabling through the tunnel, but a problem occurs when you set up an IPv6 in IPv6 tunnel: in order to set up the tunnel, you need to reach the destination of the tunnel via the unicast network. You must to be sure that the address of the tunnel end point is not included in a prefix advertised on the M6Bone if this is the case (topology choice) you need to specify a static route to the tunnel end point through the unicast network. Note that it is not possible with this solution to run multicast applications on the multicast routers IPv6 multicast routing protocols In the real world, there are many different multicast routing protocols, each with its own advantages and disadvantages. Examples of protocols: Flood and Prune Protocols (DVMRP and DM-PIM), MOSPF, Center-Based Tree (CBT, SM- PIM and BGMP). DVMRP (Distance Vector Multicast Routing Protocol) computes its own routing table to determine the best path back to the source. DM-PIM (Dense-Mode Protocol Independent Multicast) uses RPF and looks a lot like DVMRP. The most significant difference between DVMRP and DM-PIM is that PIM does not require any particular unicast protocol. For DVMRP and DM-PIM their flood-and-prune nature requires off-tree routers to keep per-source state. MOSPF is the multicast extension to OSPF (Open Shortest Path First) which is a unicast link-state routing protocol. CBT (Core-Based Trees) was the earliest centerbased tree protocol and is the simplest. The important advance that SM-PIM (Sparse-Mode Protocol Independent Multicast) made over CBT was to realize that discovering who senders are could be separate from building efficient trees from those senders to receivers. The equivalent of CBT core is calling a Rendezvous Point (RP) in PIM. The BGMP (Border Gateway Multicast Protocol) protocol is an attempt to design a true inter-domain multicast routing protocol, one that can scale to operate in the global Internet. 5. IPv6 multicast connectivity 5.1. IPv6 multicast routers However, when an IPv6 multicast routers receives a packet, it performs an RPF check. The RPF check is performing using the IPv6 unicast routing table. For this reason, it was necessary to have the same topology for IPv6 multicast and IPv6 unicast. Each router's table is unique and contains information of how and where to transmit packets to other routers. The IPv6 multicast router has to be able to manage: IPv6 in IPv6 tunnels if you have an IPv6 connectivity; IPv6 in IPv4 tunnels if you have an IPv4 connectivity; SM PIM for the construction of the shared tree; RIPng for the unicast routing (it must be possible to activate RIPng in the tunnels) 5.2. IPv6 multicast-ipv4 multicast gateway This gateway, developed by Luc Beurton from the University of South Britany, permit to have at the same and on the same session sites sending in multicast IPv4 and sites sending in multicast IPv6. In the transition between IPv6 and IPv4 it could be possible to have only one IPv6 transmission to transmit in the same time on the M6Bone (IPv6) and in the MBone (IPv4). 123

6 IPv4-IPv6 gateway IPv4 Multicast IPv6 Multicast Full multicast IPv4 (audio) Full multicast IPv4 (video) Full multicast IPv6 (audio) Full multicast IPv6 (video) An IPv6/IPv4 multicast gateway was implemented and today is use to transmit at the same time events over IPv6 and IPv4 multicast. An IPv6 Multicast/Unicast gateway was also implemented and is make it possible for a site without multicast enabled to follow any M6Bone event (like conferences, seminars, and communications between many working groups). This gateway is not installed in a definitive way on the M6Bone in spite of it has been tested with success during the broadcast of the Aristote seminar. 6. Services available on M6Bone One of the most common usages of the M6Bone is videoconferences between a large numbers of participants, without any additional configuration. Many tools are already available and can be install on almost all operating systems having the IPv6 stack. 7. IPv6 Multicast Group Management Protocol For IPv6 multicast are available the following management protocols: MLDv1 Multicast Listener Discovery Protocol (RFC2710); MLDv2 Multicast Listener Discovery Protocol (draft). References and links: DVMRP-RFC1075 Articles: [1] Mlak Mădălina <Mlak.Madalina@virgilio.it>, Multicast technology, The Sixth International Conferences on Economic Informatics, May 8-11, 2003 Bucharest, Romania [2] Tim Chown, Jérôme Durand, Pekka Savola, Stig Venås, The m6bone: International Experiments with IPv6 Multicast Recent events: Between February 2-5, 2003 at Florida International University, Miami, USA was organized an Internet2 and IPv6 workshop. 124