Scott Morris at CBTNuggets, he needs to fix his CCIE all in one video series about IPV6, he described last 32 bits are used for SNM, I had to research and find out that only 24 bits are used. I really enjoy watching and learning from Keith Barker, massive respect for your Sir, if you're listening. God bless you!
But how are multicast group members enumerated and who decides who is a valid member to receive packets for that group? Is it the router? All routers? The switches? Otherwise is it the client who decides which group traffic to listen to? It's unclear where the group memberships are "brokered" and who permits joining.
Why does PC1 not just send the packet direct to the L3 R1:G2/0 IPv6 Address configured as its default gateway? And isn't the whole point of an ARP request/Neighbor Discovery Protocol to obtain the L2 address of the destination? Won't it just have a L2 broadcast MAC when sending the Solicited Nodes Multicast anyway so it will be delivered to all the devices on that layer 2 segment? I think this example would be better with more layer 2 information. I'm just as confused as i was before i watched it...
Hello poochey1- Packets, referring to L3 are always forwarded on the local network, based on the L2 ethernet address to the next hop (router, host, etc) on that same network segment. In both IPv4 and IPv6 the nodes need to discover the layer 2 address of the next device in the path, so they can add the appropriate L2 address to the header. In IPv4 the L2 broadcast is used by ARP and processed by all local devices. In IPv6 the NDP is used and the solicitation is sent to the SLN Mcast address associated with the IPv6 address we are resolving for. There are no broadcasts in IPv6. I have several IPv6 videos on TH-cam and a complete IPv6 course on CBT Nuggets site as well for members. Hope some of that helped, Keith
Keith Barker poochey1 But what about destination MAC address of packet with Solicited-Node IPv6 Address? It still have to be MAC broadcast address so in general nothing have changed. And question still remains, why does PC1 not just send the packet direct to the L3 R1:G2/0 IPv6 Address configured as its default gateway? edit: ok, i think there is some answers - tools.ietf.org/html/rfc2464 thera are some special MAC multicast addresses too. Worth mentioning ;] An example: Let's start with IPv6 address 2001:db8::d3f5:31f8:670d. The corresponding solicited node IPv6 multicast address is ff02:0:0:0:0:1:fff8:670d. Then you apply the algorithm to get the multicast MAC address of 33:33:ff:f8:67:0d. Talking about Ipv6 without any Ethernet talk can be more confusing than informative sometimes ;)
Oskar Jagodziński If we have a multicast MAC address which will reach out our destination anyway, what's the point to use multicast on Layer 3? What about Windows machines which don't use EUI-64? ARP works fine without Layer 3 multicast or broadcast anyway.
Keith Barker Thanks for your response, Keith! Yes, I know that, but my question is why to use multicast (FF01::1:FFxx:xxxx) if we know the exact target IP address anyway (and need a response only from this particular PC)? The target host will get a multicast frame (33:33:FF:xx:xx:xx) and will process it only if there is a right destination IP - all the rest PCs will just decline it.
So this means the router is going to join that multicast group to, so it can send a message to that group so the PC can see if there was an request for the router's mac address?
But if the PC already knows it's gateway's IP address, why not just send the request to find the mac to that address instead of the solicited-node multicast address? Am I missing something? I understand how this all works just wondering if there is something I am not understanding here.
this is my biggest question haha. I still don't understand this. The way I understand it, each unicast address will have its own unique solicited node multicast address. Both unicast address and solicited node multicast address are unique. Why not just use the unicast address to get the MAC address then?
This helped me a lot Your explanation is well structured and I don't need to watch some kind of previous video to understand it. Just like Tech quckie Maybe in a live presentation you should speak a tiny bit more slowly than here but it's perfectly fine for a video
something I'm confused about is how this works with privacy extensions. Is the multicast L3 IPv6 packet, also sent to a multicast L2 ethernet address for instance? And, if so, am I to assume that the corresponding device has made a corresponding mac address to its randomly generated IPv6 privacy address?
You just said there is no layer 2 broadcast. If the frame enters the switch and is unknown, it will absolutely flood or broadcast out all ports no matter what the next protocol encapsulated is.
There is no ARP for IPv6. For resolution of a neighbor's L2 ethernet address, IPv6 uses NDP. NDP is used for many things in addition to L2 ethernet address resolution such as RS and RA as well.
![[LIVE] : ONE ลุมพินี 89 | คู่เอก "ยอดไอคิว vs คิริลล์"](http://i.ytimg.com/vi/Z3vYkHJ_5Ig/mqdefault.jpg)
Love the way he explains in every video, never getting bored.
Keith Barker has a teaching gift. His explanations are clear and his delivery is engaging. I wish my college professors were this good.
What I enjoy the most about Keith is that you can hear that he is excited to teach others and he is enjoying what he is doing. Thank you Keith!!!
Thanks for the feedback. I appreciate it.
Best wishes,
Keith
Thanks for the feedback!
Keith
Keith you're the man! This was super helpful for clarification.
Scott Morris at CBTNuggets, he needs to fix his CCIE all in one video series about IPV6, he described last 32 bits are used for SNM, I had to research and find out that only 24 bits are used. I really enjoy watching and learning from Keith Barker, massive respect for your Sir, if you're listening. God bless you!
Awesome 👏
This thing just wouldn't click in my head untill I watched the video! Thanks Keith!
Very clear, thank you Keith!!!
Many thanks for the explantation !
Thanks, very clear explanation and very positive teaching attitude!
But how are multicast group members enumerated and who decides who is a valid member to receive packets for that group? Is it the router? All routers? The switches? Otherwise is it the client who decides which group traffic to listen to?
It's unclear where the group memberships are "brokered" and who permits joining.
It helped me a lot to understand. Thanks you man!
You da man Keith!
It helped me a lot! Thanks for this explanation.
Why does PC1 not just send the packet direct to the L3 R1:G2/0 IPv6 Address configured as its default gateway?
And isn't the whole point of an ARP request/Neighbor Discovery Protocol to obtain the L2 address of the destination?
Won't it just have a L2 broadcast MAC when sending the Solicited Nodes Multicast anyway so it will be delivered to all the devices on that layer 2 segment?
I think this example would be better with more layer 2 information. I'm just as confused as i was before i watched it...
Hello poochey1- Packets, referring to L3 are always forwarded on the local network, based on the L2 ethernet address to the next hop (router, host, etc) on that same network segment. In both IPv4 and IPv6 the nodes need to discover the layer 2 address of the next device in the path, so they can add the appropriate L2 address to the header.
In IPv4 the L2 broadcast is used by ARP and processed by all local devices. In IPv6 the NDP is used and the solicitation is sent to the SLN Mcast address associated with the IPv6 address we are resolving for. There are no broadcasts in IPv6. I have several IPv6 videos on TH-cam and a complete IPv6 course on CBT Nuggets site as well for members.
Hope some of that helped,
Keith
Keith Barker poochey1 But what about destination MAC address of packet with Solicited-Node IPv6 Address? It still have to be MAC broadcast address so in general nothing have changed. And question still remains, why does PC1 not just send the packet direct to the L3 R1:G2/0 IPv6 Address configured as its default gateway?
edit: ok, i think there is some answers - tools.ietf.org/html/rfc2464 thera are some special MAC multicast addresses too. Worth mentioning ;]
An example: Let's start with IPv6 address 2001:db8::d3f5:31f8:670d. The corresponding solicited node IPv6 multicast address is ff02:0:0:0:0:1:fff8:670d. Then you apply the algorithm to get the multicast MAC address of 33:33:ff:f8:67:0d.
Talking about Ipv6 without any Ethernet talk can be more confusing than informative sometimes ;)
Oskar Jagodziński If we have a multicast MAC address which will reach out our destination anyway, what's the point to use multicast on Layer 3? What about Windows machines which don't use EUI-64? ARP works fine without Layer 3 multicast or broadcast anyway.
Boby Dome ARP only exists in IPv4. IPv6 doesn't have the ARP protocol available as part of the spec.
Keith Barker Thanks for your response, Keith! Yes, I know that, but my question is why to use multicast (FF01::1:FFxx:xxxx) if we know the exact target IP address anyway (and need a response only from this particular PC)? The target host will get a multicast frame (33:33:FF:xx:xx:xx) and will process it only if there is a right destination IP - all the rest PCs will just decline it.
So this means the router is going to join that multicast group to, so it can send a message to that group so the PC can see if there was an request for the router's mac address?
Explained well. Thank you.
But if the PC already knows it's gateway's IP address, why not just send the request to find the mac to that address instead of the solicited-node multicast address? Am I missing something? I understand how this all works just wondering if there is something I am not understanding here.
this is my biggest question haha. I still don't understand this. The way I understand it, each unicast address will have its own unique solicited node multicast address. Both unicast address and solicited node multicast address are unique. Why not just use the unicast address to get the MAC address then?
amazing explanation
Thank you! We hope that the videos here are informative!
How does PC1 get R1s Solicited Node Multicast address without knowing its MAC address?
I like the way he says Multicast :D
This helped me a lot
Your explanation is well structured and I don't need to watch some kind of previous video to understand it. Just like Tech quckie
Maybe in a live presentation you should speak a tiny bit more slowly than here but it's perfectly fine for a video
Keith why is the link local address not 128 bits why so short?
This is very helpful. Thank you so much!
something I'm confused about is how this works with privacy extensions. Is the multicast L3 IPv6 packet, also sent to a multicast L2 ethernet address for instance? And, if so, am I to assume that the corresponding device has made a corresponding mac address to its randomly generated IPv6 privacy address?
We want a reference to read about this topic
Very good explanation, thank you very much for this :)
thank you dude
Good.
good job!
You just said there is no layer 2 broadcast. If the frame enters the switch and is unknown, it will absolutely flood or broadcast out all ports no matter what the next protocol encapsulated is.
Yes that’s what my question is that’s why I watched video i thought he is going to explain that
i thought instead of ARP IPv6 uses NDP? heeelp bout to take ccent and this is now confusing
check that NDP is used for routers and switches... sowwy
There is no ARP for IPv6. For resolution of a neighbor's L2 ethernet address, IPv6 uses NDP. NDP is used for many things in addition to L2 ethernet address resolution such as RS and RA as well.