Thanks Scott =).

Glad you enjoyed it, Huzaifa. The best way to thank me is to spread the word about this content. =)

Hi There, it would be very greatfull if you explains Ping and Traceroute concept. awaiting for ur responce

@@isumitd9586 Hi Sam. In typical Networking Host B would never send an ARP to Host D. *Any* host will *only* send an ARP request to targets in *their own network*. So if Host B means to speak to Host D, Host B will ARP for the Default Gateway, not for Host D. The exception to this (and maybe this is what you are asking about) is a feature known as Proxy ARP. I wrote about that here: www.practicalnetworking.net/series/arp/proxy-arp/ But keep in mind, Proxy ARP should *not* be relied on to enable hosts in different networks to communicate. Hope this helps.

@@isumitd9586 Ahh, I see what you are asking. When the Router gets the packet, the Router will send an ARP for Host D's MAC address. It's all explained in the other video I linked.

@@isumitd9586 A Router *can* broadcast. I think you're thinking of the fact that a Router does not *forward* a broadcast. A Broadcast is simply a message sent to everyone on the network.

@@isumitd9586 The router de-encapsulates the packet to check the host destination address and then checks it’s routing table to discover the next hop address. If the host IP address is part of a connected network (which it is in this case), it will check its ARP table and if it knows the MAC address for Host D it will forward the packet. If it doesn’t know the MAC address, it will send an arp request to the known IP address of the host (and directly through all outgoing ports of any switches that may also be on the connected network between the router and the host, not true in this case). If the route to the host is via another router (next hop) in the chain, it will send an arp request to the router’s IP address for its MAC address (if not already known). The next router in the chain will check the destination IP address, check its routing table and determine where to send the packet or another arp request if it’s not already in its table. In summary: when routers receive packets they will check their ARP tables and routing tables to determine where to send them next.

Thank you, Pascal =)

Hi Neeraj, glad you enjoyed the video! I think you mean 2017 though, I wasn't making videos in 2007 =)

Whoa, what a kind comment. It's fitting, in a way, that it came from a Catalytic Centaur. Thank you for the kind words, and I'm glad you enjoyed the articles =)

Noted about STP. Thanks for the kind words, ROn. Glad you're enjoying the content!

Hi Raghavendra, glad you enjoyed it!

Most welcome! =)

You're welcome. Don't forget to check out the rest of the ARP articles: www.practicalnetworking.net/series/arp/address-resolution-protocol/

=)

Great to hear! =) So happy you're enjoying them!

Thank you, VIvek =)

Thank you. You'll probably like the "Hosts" video too: kzfaq.info/get/bejne/nb9-ZdR_lpeRp3k.html

@@PracticalNetworking I ll surely check them out

@@garvitkhurana5244 Excellent =)

Glad you enjoyed it! You might also enjoy the Networking Fundamentals series: kzfaq.info/sun/PLIFyRwBY_4bRLmKfP1KnZA6rZbRHtxmXi

Hi again. Thank you for the kind words. Glad you're enjoying this content!

Glad you enjoyed it =)

Glad you enjoy them, Karthi! I'm working on videos and new articles (pracnet.net) as we speak. Stay tuned!

Zulfikhar Amanullah Glad you liked it! I appreciate the kind words =)

Thank you, Shakar.

Noted. Thanks Don. =)

Thank you =)

Glad you enjoyed it, Mahim. You might also enjoy the full Networking Fundamentals series: kzfaq.info/sun/PLIFyRwBY_4bRLmKfP1KnZA6rZbRHtxmXi

Glad you enjoyed it. I've considered re-making a lot of my articles in video form, but its hard to prioritize that over the new content I want to create instead. =)

@@PracticalNetworking totally understand. Appreciate your time doing this. Waiting always for new videos to continue learning.

@@alonsosolano5117 Thank you for your support, Alonso =)

Thank you 😃

Glad you liked it, Zen =). You're welcome!

That's a great question! I'm glad you asked it. I actually speak to that exact question in this video: kzfaq.info/get/bejne/gtGfn8WGt9jYhKc.html The specific answer starts at 8:30~, but I'd recommend starting from the beginning to get the full context. The video will do a much better job at answering your question than I could in a comment reply.

kumaresh raj glad you enjoyed the videos! Thanks for the kind words!

It would be similar to the process outlined in this video: kzfaq.info/get/bejne/qL-fl8anzc6bkmw.html I know you watched it already, but the only difference would be the first packet would be to the DNS server to resolve the FQDN to IP mapping. Then the next packet would be an HTTP request to the IP address for the FQDN. ARP would likely only need to be completed for the first packet, as the entries will already be in the cache for the second packet.

Hi Ayan, glad you enjoyed the videos =)

Thank you kindly, SS Rot =).

Hi Ofir, That is a result of subnetting. Each hosts knows their own IP address, and their own Subnet mask. These two pieces of information can be used to determine the range of possible hosts IP addresses in their own network. Every other IP address they need to speak to is in a foreign network. You can learn about subnetting here: www.practicalnetworking.net/stand-alone/subnetting-mastery/

The procedure would be the same, except for the use of different addresses. When Host A recognizes that Host D's address is on a different network, it isn't distinguishing between a "Public" network or a "Private" IP, it is simply an IP address not in the local network. In the end, host A will never know or need to know host D's MAC address. MAC addresses are only useful for communication _within_ your network. Which means Host A only knows (and needs to know) the MAC addresses for Host B and the Router. This article will unpack it a bit: www.practicalnetworking.net/series/packet-traveling/osi-model/

Your question is a little vague. I think it may be because there are a few concepts you are unaware of. To begin, each computer understands that some IP addresses are "routable" and some are "non-routable". For example, any 127.x.x.x in this class A range is non-routable, meaning that the computer will send these packets to itself. These addresses are known as loopback addresses for this reason. Assuming that the IP address in question is routable, the computer knows that it must send data through its network interface and over an external network. Here you are again missing an important concept: subnetting. (For sake of discussion, we will consider VLANs and subnets the same thing. While fundamentally different, for the purposes of this discussion they can be said to be functionally similar.) Subnetting segregates computers into physical or logical networks known as subnets. A subnet is defined by grouping IP addresses together. While we use IP addresses in the form of AAA.BBB.CCC.DDD, our computers actually translate those "octets" into a 32-bit binary string of ones and zeroes. The leading bits in that string identify the network assignment and the remaining bits identify the host. (Imagine something like this 16-bit number 1111111100000101. The leading ones might represent a network address and the remaining bits might represent a host address. Together they form an IP address. Of course that number would actually be 32 digits long.) We can determine how many digits are in the network portion of the address by looking at the "subnet mask" assigned to the machine along with its IP address. This is where the magic happens. When a computer wants to communicate with another IP address, it uses its subnet mask to determine what network, or subnet, it is a part of. It then does the same for the IP address with which it wishes to communicate using its own subnet mask. The computer can then determine whether the two computers share the same network. So to answer your question: if the two computers are on the *same network*, it will learn the destination computer's MAC address using ARP through the process of communicating with it. If the two computers are on *different networks*, computer A never learns the MAC address of computer B because it never needs to. To sum this up simply, computers on the same network communicate by MAC address. Computers on different networks communicate by IP address only. The individual MAC addresses are only known by devices within each of their own physical (or virtual) networks.

I've created a few resources on those topics already =). Switching -- www.practicalnetworking.net/series/packet-traveling/host-to-host-through-a-switch/ Trunking / VLANS / 802.1q -- kzfaq.info/get/bejne/g9OneZSgrbXanZs.html In fact, everything I created about VLANs can be found here: pracnet.net/vlans Hope they help!

Hi Aditya, those are two different functions -- I would try not to combine them into one. =) The DORA process is DHCP, this protocol is used if a host needs an IP address. If a host doesn't need an IP address, it won't do DHCP. The ARP process resolves a L3 address to L2 address... if a host needs to resolve an address, it will do the ARP process. When ARP completes, the entry is cached for a certain amount of time, preventing the need to ARP again for the same destination for every packet. You can read more about ARP Cache timing here: www.practicalnetworking.net/series/arp/traditional-arp/

The role of L2 is hop to hop delivery. I explain that more thoroughly here: www.practicalnetworking.net/series/packet-traveling/osi-model/ The MAC address will always reflect the closest two Router or End Host MAC addresses.

Hi Donald. Each NIC has its own MAC address, it doesn't have to be "determined" by the host. A host's MAC address can be "resolved" by neighboring hosts using ARP. More details here: www.practicalnetworking.net/series/packet-traveling/osi-model/#osi-layer-2

If destination is in different network so he will send it to the gateway mac-address.

I had a discussion with a reader about this. I think it will answer your question. I'll point you that direction: www.practicalnetworking.net/series/arp/gratuitous-arp/#comment-101953

@@PracticalNetworkingI just read the comment in your article and, as I understand it, a “Gratuitous ARP” (Opcode 1) is the same “ARP Announcement” - Opcode1 - but with different names. The computer that just joined the network sends an ARP Announcement - (Gratuitous ARP) - Opcode1 - and in addition to updating the arp table of neighboring hosts, it also has its arp table updated. Updates the arp table from neighboring hosts and also receives the arp mapping from neighboring hosts.

All the information you need to build an ARP entry is in the ARP request that is broadcast to everyone on the network. So each Operating System has the choice to store or ignore ARP requests they hear that aren't targeted at them. Most OS's I've come across will ignore them. But the choice is entirely up to the operating system. Good question.

Because Host B knows the DG's IP address, but not the DG's MAC address. If it knew the MAC address, *then* it could send it Unicast straight to the Router. But since it doesn't, it must perform ARP first. (sorry to only respond just now, somehow this comment only just appeared in my feed).

Yes. It's easy for an attacker to abuse this, but they have to get on the same ip network as the intended victim (which is a little more difficult).

ARP responses are cached only temporarily. Most client's purge ARP entries around 30s or less. So at least once per minute each client will have to redo ARP queries for their neighbors. More details here: www.practicalnetworking.net/series/arp/traditional-arp/#arp-timing

That is provided "external" to the ARP process. So, for example, if I jump on host B and I type "ping 10.3.3.11", I am providing the IP address to host B directly in my command. If I am browsing the net, and I type "website.com" in a browser, DNS will convert "website.com" to 10.4.4.44 (for example), and now Host B knows it needs to send data to 10.4.4.44.

Yup! kzfaq.info/get/bejne/qL-fl8anzc6bkmw.html

Not all devices store the ARP information they hear on the wire. Mostly they just remember the ARP information they queried, or when they themselves were being queried.

@@PracticalNetworking How can I discover these devices and determine their IP Addresses?

@@machdaddy6451 On your own network? Run an ARP scan. Using a tool like nmap. =)

Typically, yes. But it doesn't have to. There can be multiple routers, and in that case the Default Gateway configured on the hosts would serve as the instructions for which router to use.

Switches do not send ARP requests. Switches are L2 devices, and don't care about L3 information (such as IP addresses). Therefore, they don't need to know any IP to MAC mapping. So my follow up question would be ... who is sending the ARP Requests?

I mean, are both same meaning

Sir, Is physical address and MAC address give same meaning?

Yep. They are different words for the same thing. =)

Because PC B only knows PC A's IP address, not it's MAC address. And the MAC address is crucial because delivery of a packet is always done using MAC addresses: www.practicalnetworking.net/series/packet-traveling/osi-model/#osi-layer-23

Anytime the Router receives an ARP Request for the IP 10.3.3.99 it will always respond with the MAC address that ends in .9999. Both Host A and Host B will use this same MAC address to send a frame to the Router so that it can be delivered to Host D (or any other host on a foreign network).

Host D knows Host B is on a different network. So Host D will arp _for the Router._ I'd recommend this video to see the whole process in action: kzfaq.info/get/bejne/qL-fl8anzc6bkmw.html

I liked your long answer you posted here: kzfaq.info/get/bejne/d85_iJWWtNbRiGQ.html&lc=UgwWra5VqLEMh2igOE14AaABAg

ok so when you refresh a site is the mac deleted and so it makes request again?

Most clients purge ARP entries every 20-30 seconds, so the ARP entry needs to be re-acquired.

@@PracticalNetworking oh i see thank you

Great question! MAC addresses are responsible for hop-to-hop delivery, and IP addresses are responsible for end-to-end delivery. You can see it clearest in this illustration: www.practicalnetworking.net/series/packet-traveling/osi-model/#osi-layer-23

Before a device gets an IP address assigned from dhcp, is Mac needed to locate the device in the network? I understand also that a network switch will save the physical port number that a Mac is communicating on, how does this work with wireless ap’s and endpoints?

​@@HAPPYSLAPS1 The MAC address is simply the address of each individual NIC. Each NIC has one prebuilt when the NIC was created. So yes, anything that connects to the Internet has a NIC and therefore already has a MAC address. It _then_ needs to acquire an IP address (using DHCP or a static configuration, etc).

This has been on my list for a while. I'd love to make a video. Hopefully soon I can schedule that in for whats next. My current/next project is organizing some live streams where I do live routing/switching labs with my viewers. That, and doing more marketing of my SSL/TLS course is taking up all my free time.

Sure. I cover that here: kzfaq.info/get/bejne/Zsd3pKp817ipYJs.html

Proxy ARP exists, certainly... but a host can *absolutely* tell when the target is on the same network or a different network. Try it for yourself, fire up Wireshark, capture the ARPs that occur when you ping a local address vs an address externally.

Thanks!