Protocol-Independent Multicast ( PIM ) is a family of multicast routing protocols for Internet Protocol (IP) networks that provide one-to-many and many-to-many data distributions over LAN , WAN or Internet. This is called protocol-independent because PIM does not include its own topology discovery mechanism, but uses routing information provided by other routing protocols. PIM does not rely on certain unicast routing protocols; it can use the unicast routing protocol used in the network. PIM does not build its own routing table. PIM uses a unicast routing table for reverse path forwarding.
There are four PIM variants:
- Sparse PIM Mode (PIM-SM) explicitly creates a unidirectional rooted root tree at the rendezvous (RP) point per group, and optionally create the shortest path tree per source. PIM-SM generally have a fairly good scale for widespread use.
- Dense PIM Mode (PIM-DM) uses solid multicast routing. Implicitly build the shortest tree by flooding the multicast traffic domain, and then trim the tree branches where there is no recipient. PIM-DM is easy to apply but generally has poor scaling properties. The first multicast routing protocol, DVMRP uses multicast-mode solid routing. See PIM Internet Standard RFC 3973.
- Two-Way PIM (Bidir-PIM) explicitly builds a shared two-way tree. It never builds the shortest path tree, so it may have longer end-to-end delays than PIM-SM, but is good-scale because it does not require source-specific circumstances. See PIM Two Way Internet Standard RFC 5015.
- Source-Specific PIM Multicast (PIM-SSM) builds trees with only one source root, offering a more secure and scalable model for a limited number of applications (mostly broadcasting content). In SSM, an IP datagram is transmitted by source S to the destination address of SSM G, and the receiver can receive this datagram by subscribing to channel (S, G). See RFC 3569 information.
PIM-SM is commonly used in IPTV systems for multicast stream routing between VLAN, Subnet or local area network.
Video Protocol Independent Multicast
Version
There are two versions of PIM. The versions are not directly compatible although they can co-exist on the same network. Network equipment can implement both versions. PIMv2 has the following fixes on PIMv1: single RPs are used per group. RP discovery is done by Bootstrap Router (BSR). The groups are a rare or dense fashion; The interface can be one of them. General improvements to the flexibility and efficiency of the protocol.
Maps Protocol Independent Multicast
Rarely mode
The Independent Multicast Protocol - Sparse-Mode ( PIM-SM ) is a protocol for efficiently routing Internet Protocol (IP) packets to multicast groups that may reach large areas and between internet domains. This protocol is named protocol-independent because it does not depend on certain unicast routing protocols for topology invention, and mode-sparse because it is suitable for very low group percentage nodes (and their routers) will subscribe to a multicast session. Unlike previous multicast multicast routing protocols such as DVMRP and multicast routing that overwhelm packets across the network and then trim the branches where there is no receiver, PIM-SM explicitly builds a tree from each sender to the receiver in a multicast group.
Multicast client
The router receives an explicit Join/Prune message from a neighboring router that has a member of the downstream group.
- To join a multicast group, G, the host passes on its membership information through the Internet Group Management Protocol (IGMP).
- The router then forwards the data packet addressed to the G multicast group only to the interface that receives the explicit join.
- A Designated Router (DR) sends a Join/Prune message periodically to a special group Rendezvous Point (RP) for each group that has an active member.
- Note that one router will automatically or statically be defined as a meeting point (RP), and all routers must be explicitly merged via RP.
- Each router along the path to RP creates wild card status for any group and sends a Merge/Prune message to the RP.
- The term route entry is used to refer to a state maintained on the router to represent a distribution tree.
- The route entry may include fields such as:
- the source address
- the group address
- the incoming interface of which packets are received
- list of outbound interfaces to which packets are sent
- timers, flag bits, etc.
- The entrance point of the wild card entrance interface leads to RP
- The outgoing interface points to a neighboring downstream router that has sent a Merge/Prune to RP message as well as a linked host that has requested membership to group G.
- This status creates a shared, RP-centered distribution tree, which reaches all members of the group.
Multicast sources
- When the first data source sends to the group, the designated router (DR) unicasts Lists the message to Rendezvous Point (RP) with the encapsulated source data packet.
- If the data rate is high, RP can send a specific Resource/Join/Prune message back to the source and the source data packets will follow the resulting forwarding status and undertake the unencapsulated trip to the RP.
- Whether they arrive encapsulated or natively, the RP forwards the de-capsulated source data packet to the RP-centered distribution tree to the group members.
- If data rates guarantee a router with a local recipient can join a specific path, shortest path, distribution tree, and trim this source package from a shared RP tree.
- For low-level data sources, neither the RP nor the last hop router need to join the shortest path tree of sources and the data packets can be sent via the shared RP-tree.
Once another router that needs to receive the group packet has subscribed, the RP will unsubscribe from the multicast group, unless it also needs to forward the packet to another router or node. In addition, the router will use reverse-path forwarding to ensure that there is no loop for forwarding packets between routers that want to receive multicast packets.
Solid mode
Multicast solid mode is one of the multicast mode that can be used to create a tree to send packets to multicast subscribers. This is the opposite of a sparse multicast.
The basic assumption behind the solid mode is that the multicast packet stream has a receiver in most locations. Modes rarely assume relatively fewer recipients. Solid mode is ideal for groups where multiple nodes subscribe to receive multicast packets, so most routers have to accept and forward these packets (high-density groups).
This distinction arises in the initial behavior and mechanisms of the two protocols. Dense Mode uses a fairly simple approach to handle IP multicast routing. The source originally broadcast to every router connected directly to it. These adjacent routers pass data to their neighbors. When the router does not want to receive this group data (if no other neighboring PIM router is present and no host is interested in the group), it sends Prune messages to indicate a lack of interest. After receiving the Prune message, the router will change its state so that it will not forward those packets out of that interface. If each interface on the router is trimmed, the router will also be trimmed.
In older IOS IOS releases, PIM-DM will overwhelm all multicast traffic every 3 minutes. This is good for low volume volumes, but not the higher bandwidth multicast packet flow. The newer Cisco IOS version supports a new feature called PIM Dense Mode State Refresh, since 12.1 (5) T. This feature uses PIM status refresh messages to refresh the Prune status of the outgoing interface. Another benefit is the recognized topology change faster. By default, PIM status refresh messages are sent every 60 seconds.
In addition, the router will use reverse-path forwarding to ensure that there is no loop for forwarding packets between routers that want to receive multicast packets. When data packets are received on a non-RPF interface, a mechanism is required to prevent loops. If the non-RPF interface is a LAN, the Assert message is sent. Non-Forwarder routers then send Prune on their RPF interface if they do not need multicast flow. Only one Prune is sent, at the time of transition to have no interface in the Outgoing Interface List (OILIST). The Prune LAN receiver delay works on it for 3 seconds, so if another LAN router still needs multicast flow, it can send PIM Join messages to disable Prune. ("The router does not need it, but I still do it!")
Suppose a router has been Pruned, and some time later the receiver requests a multicast stream with an IGMP message. The router then sends a Graft message. As a result, "hey, I need multicast streams here now".
See also
- 802.1aq Short Path Bridge
- Internet Group Management Protocol
- IP Multicast
- Multicast address
- Multicast Source Discovery Protocol
References
External links
- Gorry Fairhurst (2006). "PIM Routing" (PDF) . Archived from the original (PDF) in 2011-12-28 . Retrieved 2011-12-06 . qpimd - PIM Daemon for Quagga Ã, - Independent Protocol Multicast, previously a separate separate module, but now into an official module and provided by, Quagga Routing Suite.
Source of the article : Wikipedia
0 Comments