Created by - Orhan Ergun
Multicast PIM Dense mode vs PIM Sparse mode is one of the most important things for every Network Engineer who deploys IP Multicast on their networks. Because these two design option is completely different and the resulting impact can be very high. In this post, we will look at, which situation, which one should be used, and why. Although we will not explain PIM Dense or PIM Sparse mode in detail in this post, very briefly we will look at them and then compare them fo clarity. First of all, you should just know both PIM Dense and PIM Sparse are the PIM Deployment models. PIM Dense Mode PIM Dense mode work based on push and prune. Multicast traffic is sent everywhere in the network where you enable PIM Dense mode. This is not necessarily bad. In fact, as a network designer, we don't think there is bad technology. They have use cases If Multicast receivers are everywhere or most of the places in the network, then pushing the traffic everywhere is not a bad thing. Because when you push, you don't build a shared tree, you don't need to deal with the RP - Rendezvous Point, because Multicast Source is learned automatically. Thus, PIM Dense Mode is considered a push-based control plane and it is suitable if the Multicast receiver is distributed in most of the paces if not all, in the network. Otherwise, it can be bad from a resource consumption point of view, bandwidth, sender, and receivers process the packets unnecessarily. PIM Sparse Mode PIM Sparse Mode doesn't work based on the push model. Receivers signal the network whichever Multicast group or Source/Group they are interested in. That's why, if there is no Multicast receiver in some parts of the network, then Multicast traffic is not sent to those locations. There are 3 different deployment models of PIM Sparse Mode. PIM Sparse Mode Deployment Models PIM SSM - Source-Specific Multicast PIM ASM - Any Source Multicast PIM Bidir - Bidirectional Multicast All of these PIM Sparse mode deployment methods in the same way which Multicast receivers send join message to the Multicast Group or Multicast Source and Group. Difference between Multicast PIM Sparse Mode vs PIM Dense Mode Although technically there are so many differences, from a high-level standpoint, the biggest difference between them, PIM Dense mode works based on push-based and PIM Sparse mode works based on the Pull-based model. Multicast traffic is sent by Multicast Source to everywhere in PIM Dense mode, but Multicast traffic is sent to the locations where there are interested receivers in PIM Sparse mode. Then, we can say that, if there are few receivers, PIM Sparse mode can be more efficient from a resource usage point of view, but if there are receivers everywhere in the network, there is no problem using PIM Dense mode from a resource usage point of view.
Published - Tue, 14 Jun 2022
Created by - Orhan Ergun
Multicast BIER - RFC8279 Bit Index Explicit Replication - BIER is an architecture that provides optimal multicast forwarding through a "BIER domain" without requiring intermediate routers to maintain any multicast-related per-flow state. BIER also does not require any explicit tree-building protocol for its operation. So, it removes the need for PIM, MLDP, P2MP LSPs RSVP, etc. A multicast data packet enters a BIER domain at a "Bit-Forwarding Ingress Router" (BFIR), and leaves the BIER domain at one or more "Bit-Forwarding Egress Routers" (BFERs).The BFIR router adds a BIER header to the packet. The BIER header contains a bit-string in which each bit represents exactly one BFER to forward the packet to. The set of BFERs to which the multicast packet needs to be forwarded is expressed by setting the bits that correspond to those routers in the BIER header.Multicast BIER Advantages The obvious advantage of BIER is that there is no per-flow multicast state in the core of the network and there is no tree building protocol that sets up trees on-demand based on users joining a multicast flow. In that sense, BIER is potentially applicable to many services where Multicast is used. Many Service Providers currently investigating how BIER would be applicable to their network, what would be their migration process and which advantages they can get from BIER deployment. By using the BIER header, multicast is not sent to the nodes that do not need to receive the multicast traffic. That’s why multicast follows an optimal path within the BIER domain. Transit nodes don’t maintain the per-flow state and as it is mentioned above, no other multicast protocol is needed. BIER simplifies multicast operation as no dedicated multicast control protocol for BIER is needed while the existing protocols such as IGP (IS-IS, OSPF) or BGP can be leveraged. BIER uses a new type of forwarding lookup (Bit Index Forwarding Table). It can be implemented by software or hardware changes. Hardware upgrade requirements can be a challenge for BIER but when it is solved, BIER can be the single de-facto protocol for Multicast.
Published - Wed, 25 May 2022
Created by - Orhan Ergun
Multicast PIM SSM - Source Specific Multicast from a design point of view will be explained in this post. The Shortest Path Tree concept, Advantages, and disadvantages of Multicast PIM SSM will be covered as well. What is Source Specific Multicast - PIM SSM? PIM is a Multicast Routing Protocol. There are two categories of PIM protocol. PIM Dense mode and PIM Sparse Mode. PIM Sparse Mode has 3 different modes of deployment. PIM SSM - Source Specific Multicast, PIM ASM - Any Source Multicast, and PIM Bidir - Bidirectional Multicast. In this post, we will only cover PIM SSM but for the other PIM Sparse mode and PIM Dense mode design and deployment posts, place check Multicast category. PIM SSM is called Source-Specific because Multicast receivers not only specify the Multicast Group that they are interested in but also they can signal to the network which course they are interested in or they are not interested in. PIM SSM in the Routing Table In the routers, we have multicast routing tables. SSM Multicast routing entries in the routers are seen as S, G. S stands for multicast Source and G is used for multicast Group. Source information has to be known in PIM SSM. So, Source Specific Multicast requires Multicast Receivers to know who is the Multicast Source for the Multicast Group/Groups. In PIM ASM - Any Source Multicast, for example, routing entries in MRIB (Multicast Routing Table) are seen as *, G, because Source information is not known with PIM ASM. PIM SSM Range by IANA IANA reserved 232/8 for PIM SSM usage. Although if you don't configure the 232/8 address range for Source Specific Multicast usage, SSM still works. But it is good to use this range for better troubleshooting. When the operator sees this range, they can immediately identify that the information is from the SSM range. Last but not least, SSM requires IGMPv3. So, if Source Specific Multicast will be used, Multicast receivers need to support IGMPv3. If somehow receivers don't support IGMPV3, then at the Multicast last-hop router, IGMPv2 to v3 mapping can be done. It is called SSM Mapping. Where PIM SSM Should be used? Because Source Specific Multicast - PIM SSM comes with more specific information (Not just group, but the source and the group), it is good for optimal routing. Let me explain this point a little bit more. SSM uses the Shortest path tree. The shortest-path tree, which is also known as the Source-based tree is using the IGP shortest path between the Multicast Sender and the Receiver. Having SPT (Shortest Path Tree), and using the shortest IGP path, means Optimal Routing basically. Optimal Multicast Routing. Let's have a look at Shortest Path Tree. In the above topology, let's assume all the interface costs are the same. So, clearly, Sender 1 and Sender 2, are using the IGP shortest path to reach Receiver 1 and Receiver 2. Sender 1 has two paths for example to reach Receiver 2.. 1-2-3-5 and 1-5. Because SPT (Shortest Path Tree) uses the shortest IGP cost, 1-5 path is used to send Multicast traffic. Same thing can be said for the Unicast routing as well. Whenever there is a more specific entry in Unicast, you can have Optimal Routing. Whenever there is summarization, it means, you will have less entry in the control plane, which can increase the sub-optimal routing chance. Thus, if we want to increase the Optimal Routing chance, we would like to use PIM SSM - Source Specific Multicast. But, what can be the tradeoffs using SSM - Source Specific Multicast, where it shouldn't be used? Where PIM SSM Shouldn't be used? PIM SSM, because specifies each and every source with S, G entries, in the control plane, might create a scalability issue. So, if the device resources might be an issue, if you have low-end routers in your network, using PIM SSM - Source Specific Multicast may not be your option. Also, more information in the control plane might extend the troubleshooting time when there is a problem in the network. Last but not least, because with PIM SSM, we have more specific Multicast routing table entries, compare to PIM ASM and PIM Bidir, more entries might lead to a longer network convergence time in the case of failure.
Published - Mon, 11 Apr 2022
Created by - Orhan Ergun
In this blog post, I will explain some of the Multicast basics that most of us look for. MPLS Multicast and many other Multicast Design, Troubleshooting, and Multicast Deployment topics are explained in the different blog posts on the website. Also, this post will cover the many fundamental Multicast frequently asked questions briefly. For a more detailed explanation of the particular topic, you can check our other blog posts on the website. Before we start, please note that if you are looking for IP and MPLS Multicast video course, you can click here. What is Multicast used for? There are many reasons in the real life for Multicast, but mostly we are seeing it in the financial networks, stock exchange, Large Campus Networks for IP Surveillance, and IPTV Multicast purposes. When it comes to the deployment details, although we will cover them in separate blog posts, in IPTV, Source Specific Multicast, in Financial Networks, Bidirectional Multicast is used. Also, using Multicasting provides resource optimization, which means, less bandwidth, less source, and receiver CPU and Memory usage it can provide. IP Multicast RoutingThere are many Multicast Protocols for Multicast to work in the Networks but when it comes to Multicast Routing, today, as of 2022, PIM is the only Multicast Routing Protocol. PIM stands for Protocol Independent Multicast and here I covered the details for it. In early 2000, DVMRP was used before PIM. DVMRP stands for Distance Vector Multicast Routing Protocol and it is similar to PIM Dense Mode. We cover PIM Dense and PIM Sparse Mode, as different PIM deployment models in a different blog post on the website. IP Multicast Routing Protocol, PIM is used for providing Multicast Streaming from Multicast source to Multicast Receiver, by using Layer 3 IP Unicast Infrastructure. What is Multicast Traffic?We can carry video, voice, or data multicast traffic between the source and the receiver. All different data types can be Multicast Traffic. Important to understand that, we are sending the same information to multiple receivers, at the same time and Network devices replicate this information. What is Multicast IP Address?Many people confuse this. We have 3 different entities to that we assign an IP address. Multicast Source, which is also referred to as Multicast Sender, Multicast Receiver, and Multicast Group. IP Multicast Source and IP Multicast Receiver always get Unicast IP Address. We only assign Multicast IP Addresses to the Multicast Groups. Okay, then the question is below. What is Multicast IP Address Range? Multicast IP Address range is also known as Class D IP Address Range and it is 224/4. It starts from 126.96.36.199 and up to 188.8.131.52 Quite a big range and inside this, IANA reserved 232/8 for Source-Specific Multicast, SSM purpose. Which is Multicast MAC Address?The multicast MAC address is a special value that begins with 01-00-5E in hexadecimal. The remaining portion of the Multicast MAC address is created by converting the lower 23 bits of the IP multicast group address into 6 hexadecimal characters Between Multicast IP Address and multicast MAC Address, there are 5 Bits overlaps. Thus, 32 different Multicast groups can share the same Multicast MAC address. This usually is a consideration when we do the planning for Multicast IP addresses for the Multicast Groups. Which of the following is a multicast routing protocol? Let's make a quick test. PIM HSRP BFD NHRP The answer is PIM. HSRP is a first-hop redundancy protocol. BFD is used for failure detection for fast convergence. NHRP is the Next Hop resolution protocol, used in DMVPN. In this post, only a few of the multicast basic topics are mentioned briefly, for a deeper explanation of many other Multicast topics, please check other posts on the website.
Published - Wed, 30 Mar 2022
Created by - Orhan Ergun
BIER is Bit Indexed Explicit Replication which is the newest proposal for IP Multicast. Although I say IP Multicast, of course, it works on MPLS networks as well. You can take our IP and MPLS Multicast Training for more detail. BIER works by assigning every edge device a Bit Mask position. Then, instead of sending Multicast packet to each destination IP address (Receiver IP address), basically, it sets the Bit positions and saves the amount of data plane state. It uses Unicast transport as underlay reachability, and Bit Mask is advertised through the IGP control plane. So, OSPF and IS-IS newly assigned TLVs to handle the BitMask to Edge device (BFER - Bit Forwarding Edge Router in BIER terminology) assignment and distribution. It is in theory can be used not only for multicast but also for Unicast traffic as well. When we use it, we don't need to have mLDP, RSVP P2MP LSPs, or PIM in the Core Network (Of course at the Edge, you can still have towards the customer in mVPN scenarios). So basically, by removing those protocols from the network, in theory, the simpler network design you should have. I am saying in theory, because having less protocol doesn't always mean, having a simpler design. Because we would be throwing the complexity to the protocol in this case. So carrying Bit mask positions in OSPF and IS-IS makes these protocols codes more complex. So, there is no free lunch in network design. I recorded a very long, detailed video about BIER with the two guys who actively participated for the invention of this protocol. Tony P has been there since day 1 of this protocol and hopes you will find our video useful if you are reading this post. [embed]https://www.youtube.com/watch?v=p7bETttpPh4&t=22s[/embed] In this video, Orhan Ergun, Dr. Tony Przygienda, and Jeff Tantsura discuss a pretty new Multicast Architecture - BIER. BIER is radically different than traditional Multicast.BIER uses unicast transport to provide Multicast without Ingress Replication! - So Scalable. (It keeps state at the Edge too, not in the Core as traditional GRE, mLDP or RSVP solutions in the Core) Scalability, Simplicity, Agility, are the pros of BIER, lack of large-scale deployments, hardware change/Silicon Support current disadvantages. But many Core networks are considering it and I want my followers to be aware of BIER!
Published - Thu, 27 Jan 2022