03-02-2023 09:12 AM - edited 03-02-2023 09:50 AM
Hello.
2 workstations, a switch, then a router live on the same subnet. Both workstations have the same router as their default gateway. A layer 2 switch lives between the router and the workstations. All devices have correct ARP and/or mac-address tables. There exist no VLANS.
Workstation1 communicates to Workstation2.
My understanding is that...
1. The packet path would be workstation1 <===> switch <==> workstation2.
2. The workstation1 egress destination mac-address would be the mac-address of workstation2.
3. The switch would change nothing, but merely move the frame from the port connected to workstation1 to the port connected to workstation2.
4. Workstation2 would receive the packet, and then process the packet's data through OSI level 7.
May you please confirm this is correct?
Thank you.
Solved! Go to Solution.
03-02-2023 10:32 AM
"May you please confirm this is correct?"
More or less it is. Much depends how technical you want to get. I.e. 30,000 foot view or down in the mud.
For example, #1 starts with "packet path", L3; #2 discusses MAC addresses, L2; #3 discusses frames, L2, and "merely", often quite a bit involved under the covers for that "merely"; and #4 we're back using "packet", and passing packet's data through OSI L7, when the OSI model might not be, and probably not, used at all (e.g. TCP/IP model), or OSI model only partially implemented. (NB: the OSI model is a nice logical way of abstracting networking, but actually networking implementations aren't bound by it.)
Or, for another example, on #2, you appear to be assuming you're using unicast MACs, but assuming we're discussing Ethernet, there's also multicast and broadcast MACs, which would also support communication between two hosts. Or, you note "mac-address of workstation2", yea, but a WS can have multiple NICs each (I believe) with multiple MACs.
Don't misunderstand, as a high level understand, it's okay, you're correct, but networking rabbit holes can run surprisingly deep. One can wonder about some of your preconditions, such as there's a router on the same subnet (yea and so? [you're trying to confirm it's not involved?]), there's no VLANs (again, an so? [VLANs don't matter, for your description, same L2 broadcast domain does]). Or, devices are on same subnet (and so? [your situation can work with the two WSs on different networks, w/o using router]). Again, networking rabbit holes can run deep. ; )
Anyway, I write the above, to hopefully have you understand, your understanding is correct at a high level, which is fine. Just realize as you get closer to the mud (like troubleshooting a complex L2 issue), high level understanding can start to fall apart.
BTW, as you continue to learn, we don't throw away high level understandings, as we often use whatever level of understanding is needed in any particular situation.
Hope I have not confused you. Feel free to post any follow on questions.
03-02-2023 09:24 AM
I believe this was asked before and google has many examples of how that works.
below video give you directions:
03-02-2023 09:27 AM - edited 03-02-2023 11:21 AM
follow........
03-02-2023 09:29 AM - edited 03-02-2023 09:49 AM
Thank you, Balaji, for your reply.
That is a helpful video, but in its scenario there doesn't exist a router on the subnet, or workstations with DG's configured in their routing tables.
I expect my originally described understanding is correct. Is it correct?
Thank you.
03-03-2023 12:10 AM
Near by, @Joseph W. Doherty beat me with goog explanation (he does always though).
03-02-2023 10:32 AM
"May you please confirm this is correct?"
More or less it is. Much depends how technical you want to get. I.e. 30,000 foot view or down in the mud.
For example, #1 starts with "packet path", L3; #2 discusses MAC addresses, L2; #3 discusses frames, L2, and "merely", often quite a bit involved under the covers for that "merely"; and #4 we're back using "packet", and passing packet's data through OSI L7, when the OSI model might not be, and probably not, used at all (e.g. TCP/IP model), or OSI model only partially implemented. (NB: the OSI model is a nice logical way of abstracting networking, but actually networking implementations aren't bound by it.)
Or, for another example, on #2, you appear to be assuming you're using unicast MACs, but assuming we're discussing Ethernet, there's also multicast and broadcast MACs, which would also support communication between two hosts. Or, you note "mac-address of workstation2", yea, but a WS can have multiple NICs each (I believe) with multiple MACs.
Don't misunderstand, as a high level understand, it's okay, you're correct, but networking rabbit holes can run surprisingly deep. One can wonder about some of your preconditions, such as there's a router on the same subnet (yea and so? [you're trying to confirm it's not involved?]), there's no VLANs (again, an so? [VLANs don't matter, for your description, same L2 broadcast domain does]). Or, devices are on same subnet (and so? [your situation can work with the two WSs on different networks, w/o using router]). Again, networking rabbit holes can run deep. ; )
Anyway, I write the above, to hopefully have you understand, your understanding is correct at a high level, which is fine. Just realize as you get closer to the mud (like troubleshooting a complex L2 issue), high level understanding can start to fall apart.
BTW, as you continue to learn, we don't throw away high level understandings, as we often use whatever level of understanding is needed in any particular situation.
Hope I have not confused you. Feel free to post any follow on questions.
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: