MAC address
_ahmed
Posted messages
42
Status
Membre
-
brupala Posted messages 111985 Registration date Status Membre Last intervention -
brupala Posted messages 111985 Registration date Status Membre Last intervention -
1- What are the types of MAC addresses to which a station must respond?
2- A multicast group is identified by @ 224.16.21.5. Knowing that the range of @ multicast defined for encapsulation of @ multicast IP addresses is: 01-00-5E-00-00-00 to 01-00-5E-7F-FF-FF
.
To what MAC @ will the Ethernet adapter with unicast @ 08-00-02-2D-75-BD located on the subnet 10.10.31.0/24 respond?
2- A multicast group is identified by @ 224.16.21.5. Knowing that the range of @ multicast defined for encapsulation of @ multicast IP addresses is: 01-00-5E-00-00-00 to 01-00-5E-7F-FF-FF
.
To what MAC @ will the Ethernet adapter with unicast @ 08-00-02-2D-75-BD located on the subnet 10.10.31.0/24 respond?
Configuration: Windows XP Internet Explorer 6.0
6 réponses
In fact, for multicasts, there is a relationship between IP addresses and MAC addresses.
Multicast IPs range from 224.0.0.0 to 239.255.255.255. So if a host sends a multicast packet to a given address, say 224.0.0.10, the switch (which only handles MAC addresses) must be able to send frames only to the multicast recipients.
To achieve this, there is a convention that states that multicast MAC addresses start with 01-00-5E, and the remaining 3 bytes are derived from the rightmost 23 bits of the multicast IP address.
For IP 224.0.0.10, the MAC address will be 01-00-5E-00-00-0A.
Multicast IPs range from 224.0.0.0 to 239.255.255.255. So if a host sends a multicast packet to a given address, say 224.0.0.10, the switch (which only handles MAC addresses) must be able to send frames only to the multicast recipients.
To achieve this, there is a convention that states that multicast MAC addresses start with 01-00-5E, and the remaining 3 bytes are derived from the rightmost 23 bits of the multicast IP address.
For IP 224.0.0.10, the MAC address will be 01-00-5E-00-00-0A.
Imanova
Thank you for the information.
hi,
for MAC addresses, the broadcast is: ff-ff-ff-ff-ff-ff and this applies to any network or subnet!!!
for MAC addresses, the broadcast is: ff-ff-ff-ff-ff-ff and this applies to any network or subnet!!!
1- What are the types of MAC addresses to which a station must respond?
2- A multicast group is identified by @ 224.16.21.5. Knowing that the range of multicast addresses defined for the encapsulation of multicast IP addresses is: 01-00-5E-00-00-00 to 01-00-5E-7F-FF-FF
.
To which MAC address will the Ethernet adapter with unicast address 08-00-02-2D-75-BD, located on the subnet 10.10.31.0/24, respond?
2- A multicast group is identified by @ 224.16.21.5. Knowing that the range of multicast addresses defined for the encapsulation of multicast IP addresses is: 01-00-5E-00-00-00 to 01-00-5E-7F-FF-FF
.
To which MAC address will the Ethernet adapter with unicast address 08-00-02-2D-75-BD, located on the subnet 10.10.31.0/24, respond?
What is this story of mixing MAC address and IP address in the same question?
There is no correlation between the two in IPv4.
First: a station must respond
a station must not respond at the MAC level, it is the different applications that respond, not layer 2.
However, it can only receive (in non-promiscuous mode) broadcast, multicast, and unicast frames intended for its own MAC address.
--
There is no correlation between the two in IPv4.
First: a station must respond
a station must not respond at the MAC level, it is the different applications that respond, not layer 2.
However, it can only receive (in non-promiscuous mode) broadcast, multicast, and unicast frames intended for its own MAC address.
--
and ... Voili Voilou Voila !
If I understand correctly what is explained in "Interconnections Second Edition" by Radia Perlman, chapter 15.2.3: Mapping NL Multicast to DL Multicast, then an Ethernet card can listen to what is addressed to its actual MAC address as well as to what is addressed to the MAC address corresponding to the multicast group (here inferred from the multicast IP address 224.16.21.5).
What role does the IP address of the network to which this Ethernet card is connected (10.10.31.0/24) play? I have no idea. I tend to think that it is a superfluous piece of information, but it has the merit of fully specifying the issue...
If someone can confirm this based on an RFC (perhaps the one from IGMP), I think that would settle the question.
What role does the IP address of the network to which this Ethernet card is connected (10.10.31.0/24) play? I have no idea. I tend to think that it is a superfluous piece of information, but it has the merit of fully specifying the issue...
If someone can confirm this based on an RFC (perhaps the one from IGMP), I think that would settle the question.
Which MAC address will the Ethernet adapter reply to for the unicast 08-00-02-2D-75-BD located on the subnet 10.10.31.0/24? What is the relationship between the MAC address and the IP address? I really don't understand, you're mixing things up here :s
>I really don't understand, you're mixing things up here :s
You need to tell that to Mr. IETF!
I'll try to clarify, stop me if I'm wrong.
The basic problem is: In a LAN, what should be the destination MAC address used if the source packet is a multicast IP?
- One solution would be, I think, for the LAN router to know all the MAC addresses interested in this multicast and to duplicate the "frame" as many times as necessary: 1 IP packet arrives (regardless of layer 2) and multiple frames are transmitted.
- Another solution is to use a multicast MAC address that is automatically calculated from the multicast IP address. For cost reasons, the IETF decided that this mapping would not be perfect; a multicast MAC address actually corresponds to 32 potential multicast IP addresses. (so it's not a one-to-one mapping :-)
In this scenario, an IP packet arrives and one frame is transmitted.
Nowadays, what is the interest of such a method since connections are mainly point-to-point (indeed, duplication occurs in any case), I ask you the question...
I even wonder why layer 2?
You need to tell that to Mr. IETF!
I'll try to clarify, stop me if I'm wrong.
The basic problem is: In a LAN, what should be the destination MAC address used if the source packet is a multicast IP?
- One solution would be, I think, for the LAN router to know all the MAC addresses interested in this multicast and to duplicate the "frame" as many times as necessary: 1 IP packet arrives (regardless of layer 2) and multiple frames are transmitted.
- Another solution is to use a multicast MAC address that is automatically calculated from the multicast IP address. For cost reasons, the IETF decided that this mapping would not be perfect; a multicast MAC address actually corresponds to 32 potential multicast IP addresses. (so it's not a one-to-one mapping :-)
In this scenario, an IP packet arrives and one frame is transmitted.
Nowadays, what is the interest of such a method since connections are mainly point-to-point (indeed, duplication occurs in any case), I ask you the question...
I even wonder why layer 2?
Off topic, but it has the property of annoying me:
why don't you say "the basic problem" instead of "the basic issue"?
the source IP packet doesn't matter much, what is important is the destination MAC, if it's a multicast, everyone receives it, that's how it is.
the transposition of IP multicast / ethernet multicast is another subject.
otherwise, indeed,
multicast is a bastard technique inherited from hubs and adapted more or less well, it's better to duplicate the frame and it's true that in the long run, multipoint networks will disappear to make way for point-to-point networks where indeed the local address doesn't matter at all... like X25 (lapB).
--
and ... there you go!
why don't you say "the basic problem" instead of "the basic issue"?
the source IP packet doesn't matter much, what is important is the destination MAC, if it's a multicast, everyone receives it, that's how it is.
the transposition of IP multicast / ethernet multicast is another subject.
otherwise, indeed,
multicast is a bastard technique inherited from hubs and adapted more or less well, it's better to duplicate the frame and it's true that in the long run, multipoint networks will disappear to make way for point-to-point networks where indeed the local address doesn't matter at all... like X25 (lapB).
--
and ... there you go!
Hello,
dear CCM members, I am a computer science student and I am preparing my final project. I am working on mesh wireless networks and I want to add multicast to a Layer 2 protocol. So how can I indicate in my frame that it is a broadcast address? I know that MAC addresses are static (unique: manufacturer). Please explain to me.
The question: Are there virtual (calculable) MAC addresses that define Layer 2 multicast? Thank you in advance.
dear CCM members, I am a computer science student and I am preparing my final project. I am working on mesh wireless networks and I want to add multicast to a Layer 2 protocol. So how can I indicate in my frame that it is a broadcast address? I know that MAC addresses are static (unique: manufacturer). Please explain to me.
The question: Are there virtual (calculable) MAC addresses that define Layer 2 multicast? Thank you in advance.
Hello,
on a Mac address, you must first absolutely set the I/G bit for multicast:
the first destination octet will then have an odd value: this is the characteristic of a broadcast/multicast.
In fact, in practice, it will always be OUI 01-00-5E for the first 3 octets
then it is a matter of defining the multicast group, that's more complicated, but not very useful, as it is rechecked at layer 3.
see also:
http://www.tcpipguide.com/free/t_TCPIPAddressResolutionForIPMulticastAddresses.htm
P.S.,
you could have asked your question in a new discussion
on a Mac address, you must first absolutely set the I/G bit for multicast:
the first destination octet will then have an odd value: this is the characteristic of a broadcast/multicast.
In fact, in practice, it will always be OUI 01-00-5E for the first 3 octets
then it is a matter of defining the multicast group, that's more complicated, but not very useful, as it is rechecked at layer 3.
see also:
http://www.tcpipguide.com/free/t_TCPIPAddressResolutionForIPMulticastAddresses.htm
P.S.,
you could have asked your question in a new discussion
Here,
funnily enough,
I’m taking advantage of the return to this discussion from antiquity to add a little update:
Today, IPv4 is going into the past and we must evolve to IPv6, whether we like it or not.
The destination MAC address for IPv4 multicast Ethernet was 01-00-5E-xx-xx-xx
the XXs representing the multicast group at the IPv4 level
In IPv6, the destination MAC address becomes 33-33-xx-xx-xx-xx, with the XXs still representing the multicast group; we see that the differentiation capability of the group at the MAC level has increased by 8 bits (9 in fact), thus multiplying the number of groups to manage for switches by 512.
If we relate this to IPv6, there are 112 bits that represent the group, whereas in IPv4, we actually only had 28 for the same numbering; it is impossible to manage switching tables in terms of the number of groups.
In fact,
and I regret this,
IPv6 makes much more use of multicast than IPv4; already, ARP (broadcast) has disappeared along with all broadcasts and has been replaced by Neighbor Solicitation, which is a multicast to a group represented by the last three bytes of the IPv6 address to which we request its MAC address.
Each machine is required to listen to this multicast group corresponding to its IPv6 address(es).
funnily enough,
I’m taking advantage of the return to this discussion from antiquity to add a little update:
Today, IPv4 is going into the past and we must evolve to IPv6, whether we like it or not.
The destination MAC address for IPv4 multicast Ethernet was 01-00-5E-xx-xx-xx
the XXs representing the multicast group at the IPv4 level
In IPv6, the destination MAC address becomes 33-33-xx-xx-xx-xx, with the XXs still representing the multicast group; we see that the differentiation capability of the group at the MAC level has increased by 8 bits (9 in fact), thus multiplying the number of groups to manage for switches by 512.
If we relate this to IPv6, there are 112 bits that represent the group, whereas in IPv4, we actually only had 28 for the same numbering; it is impossible to manage switching tables in terms of the number of groups.
In fact,
and I regret this,
IPv6 makes much more use of multicast than IPv4; already, ARP (broadcast) has disappeared along with all broadcasts and has been replaced by Neighbor Solicitation, which is a multicast to a group represented by the last three bytes of the IPv6 address to which we request its MAC address.
Each machine is required to listen to this multicast group corresponding to its IPv6 address(es).