Hello Bilal,
The term “group address” might be a bit confusing. In reality, a multicast address is just a destination IP address. The main difference is that we call them “group” addresses since a single multicast IP address can represent a “group” of receivers.
We still have one destination address (a multicast address) but instead of forwarding an IP packet from A > B, we can forward it from A > everyone who listens to the multicast address.
Rene
Hi Rene,
could you please write a topic about multicast sparse mode between Vlans SVI at on core switch ( layer 3.
how the multicast routing table will be one we enabled that in one Vlan then once add one more vlan in the multicast group .
regards
Hello ABC
You can make suggestions about topics at this page:
As for your specific question, you shouldn’t see a difference between how a multicast routing table will be for a L3 switch with multiple SVIs or just a router with multiple ports as members of a single multicast group. You can actually modify Rene’s topology in this lesson to accommodate that and see it first hand.
I hope this has been helpful!
Laz
Rene,
Can I use a layer 3 switch instead of a router in the labs for multicast?
Hello David
If the switch platform supports multicast functionality such as PIM and RP and the rest, then sure, you can use an L3 switch. The Catalyst 3850 with the appropriate IOS for example, supports all of the required parameters.
For features such as IGMP, an L3 switch would also do the trick since it does support L2 multicast features as well.
I hope this has been helpful!
Laz
Hi Laz and Rene,
I might have known the answer for my next question before, but right now after memorizing this topic again, I can’t find a solution for the behaviour as described below:
We can overcome duplicate multicast traffic by using the RPF check to the (S,G) group, and the router that receive the duplicate streams will only accept the packets from the lower metric path - which means only the best path link will accept the multicast traffic if it will be duplicated/received from different paths. (I’ts not the same topology as with PIM Assert, I’m talking about different physical links which means different segments to receive the traffic from)
What happen when the traffic will come from different paths with the same metric?
How would the duplicate data can be avoided in such scenario?
(I tried to lab such a scenario and forced igmp join from 2 different links for the same group , I can see the join messages but for some reason, one of the routers in the middle between the RP wouldn’t insert the interface which received the igmp report message to the OIL - and I can’t figure out why this is happening, after all both of the two middle routers do advertises the IGMP REPORT as a PIM JOIN but one of them doesn’t insert the port from which the igmp report came from to the OIL for that multicast group)
Thanks you vey much again!
Hello Nitay
By default, when you have multiple equal-cost paths in multicast, the PIM neighbor with the highest IP address will be used.
You can find out more about this behaviour at the following Cisco documentation.
I hope this has been helpful!
Laz
I have a question about the Multicast group creation and assignment. How do I know what group to make? I am trying to clean up a botched Multicast issue configuration and I am still pretty new to multicasting and your material has been great. But I have multiple groups showing up on my L3/L2 switches but no interfaces are assigned to a multicast group or an RP. I see you use 239.1.1.1, but I have 224.0.1.60, 239.1.0.1, 239.255.255.250/253/254 and 225.86.67.83, how do I know what group to create?
Hello Patrick
The multicast addresses that you can use are in the range of 224.0.0.0 to 239.255.255.255, but there are some reserved addresses within that range. The IP addresses you are using are all within the free for use ranges, except for 224.0.1.60 which is a well known multicast address registered as hp-device-disc as seen from the IANA address. Without researching it further, it is likely used by a service on an HP network device.
Now having said that, how do you know what groups to create? It depends on what you want to achieve. Remember that the groups are actually created by the hosts trying to join a multicast group. It is the ip igmp join-group command (in the case of a Cisco device) that causes a host to begin sending join requests, which eventually adds the multicast group to the routing table of the routers. (In the case of a computer, the OS is responsible for sending out these join requests based on some application that is managing these communications.)
So what groups will be created simply depends on the hosts (and the multicast sources) and the applications that they are running.
I hope this has been helpful!
Laz
Hi Rene,
Thanks for this Multicast lesson, this was very helpful to understand the basics.
Laszlo
Hello Laz and Rene,
Thank you for the lesson, it’s very clear but I have two questions 
I understand that pim sparse mode is preferred but If we have only one router with several vlans in the network; does it matter if we have pim dense mode or sparse mode?
My second question is related with igmp snooping querior which I didn’t understand much. I read on a document that it can be used instead of multicast router. let’s say if we have a video source in local Lan in a vlan 10 and clients reside in vlan 10 and vlan 20; the router is an old one which can’t handle multicast routing or if we don’t have access to the router; having one switch (L2orL3?) configred as igmp snooping querior per vlan; will help the hosts to get the multicast stream?
Thanks for your valuable work.
Hello Ike
If you only have a single router, then it won’t make a difference if you are using dense or sparse mode. These modes affect how routers forward multicast traffic to other routers. If you have only a single router, then the mode won’t affect how multicast functions.
Concerning the use of a multicast querier, you can take a look at the following lesson which describes how to implement IGMP snooping with and without a router.
I hope this has been helpful!
Laz
Thank you very much Laz, that is what I was looking for.
Hi, new user here… just wondering as this topic has many options…
I’m trying to set up Windows Deployments with Multicasting… We have a few HP L3 Switches, one core that performs routing, the rest are departmental access switches… if the WDS server is on vlan 10 on Switch1 and the receivers are on vlan 20 - Switch 3, do I need PIM setup to route between vlans? do I also need RIP… if you can shed some light on this would be appreciated… no routers involved just Switches… ty
Hello Yosi
From my understanding, you have a single L3 switch as the only routing device in your topology. In such a case, as long as multicast routing is enabled on this device, it should, by default, route multicast traffic. It will, by default, use dense mode, and will choose itself as the querier, but that’s OK since it is only a single router.
In order to reach other networks (such as the Internet), you simply need to set up some routing to send traffic outside of your network, but I assume that doesn’t include any multicast traffic, so you’re still OK. Such routing can be applied either statically or using any routing protocol, but again that doesn’t affect your multicast topology.
Does that make sense?
I hope this has been helpful!
Laz
Hello!
have a big question)
can one group have multiple sources with different source ip (same traffic)?
The question is in the context of pim ASM only.
How is this supposed to work? How does SM RP choose which source to send traffic from when using pim?
And how does it work with PIM DM?
Hello Georgiy
The quick answer is yes. Remember, that Any-Source Multicast (ASM) means that the multicast mechanism is creating groups of destinations that correspond to specific multicast addresses. You can have multicast traffic from multiple sources going to the same multicast group.
For example, you may have a multicast group in your network of 239.1.1.1 to which 12 hosts on your network belong. Any device can send traffic to this destination, and that traffic will reach all 12 hosts. So you can have 5 servers for example (multiple sources) sending data of any kind to this group, and the traffic would reach the members of that group.
When we’re talking about ASM, the multicast groups created are groups of destinations. Any source can use those multicast destinations, that’s part of how ASM works.
The SM RP does not choose which source to send traffic from. The SM RP receives traffic from any source that chooses to send traffic to a multicast group. The choice is the source’s and not the RP’s.
There are no special considerations needed for either DM or SM when you have multiple sources since this is the normal behavior of such a configuration. Only with source-specific multicast do we care what the source is, and this is regulated at the host itself, and not by some other centralized service like an RP.
I hope this has been helpful!
Laz
that is, traffic to hosts from all servers at the same time, right?
And then it is the client’s concern from which source to take?
Very strange … Let’s say there are 5 sources that broadcast video in the same group 239.1.1.1 one of the clients launched vlc and wants to watch this video and rp sends traffic to it from all servers … as I understand it, it’s not possible to watch normally since the traffic will be heavily duplicated … With ASM, the host does not initially determine the source, it is logical to assume that in the future it will not be able to separate one from the other
Hello Georgiy
Sure, why not? Any host can send anything it likes to whatever destination IP it likes, however it likes on a network, whether it is unicast or multicast.
According to how SSM works, yes. If you look at the SSM Lesson, the actual “source specific” configuration takes place in the client itself with the commands:
R3(config)#interface fastEthernet 0/0
R3(config-if)#ip igmp join-group 232.1.1.1 source 192.168.12.1
Hmm… First of all we need to clarify a few things. SSM does not use an RP like PIM sparse mode does. However, what does take place is the SSM multicast client sends an IGMP message to the upstream router to let it know that it accepts multicast traffic only from a particular source. So if another source tries to send traffic to that multicast group, the upstream router will not send that multicast traffic to the particular host. (This is clearly described in the SSM lesson).
So in your scenario, let’s say there are 5 sources that send traffic to 239.1.1.1, and a particular host is configured to accept multicast traffic only from a single source. The upstream router will only deliver multicast traffic to the host that comes from the configured source address.
Why would traffic be duplicated? Each of those 5 servers is delivering something different. Multicast sources are independent of one another. Are you suggesting that these 5 sources are actually working together (in some sort of redundant arrangement) to deliver a single stream of multicast data? Can you clarify the setup you are describing here?
Remember that multicast does not operate alone. When you have multicast services, such as a live video stream, there are applications that are running on computers that will allow clients to communicate with sources to begin and end streaming. Clients via applications like VLC will be able to communicate with the source and request or end service. These are communications that take place at higher layers (application layer) above the multicast routing that enable the negotiation of what you actually want to do. Even if there are 5 multicast streams coming into a particular computer, the IP addresses and transport layer port numbers will allow the PC to differentiate between the streams of data and associate them with the correct applications running on the device. Does that make sense?
I hope this has been helpful!
Laz
Hello,
In the previous lessons we didn’t talk about IGMP join but we said IGMP Membership Report. Is that the same thing?
Thanks,