Does Tagging/Encapsulation always occur?

up2thetime

Does Tagging/Encapsulation always occur?

If two switches are connected together with no VLANs configured on each (Switch 1 and Switch 2 just have VLAN1 configured by default with all the switchports contained in VLAN1). If these switches are connected, will 802.1Q tagging take place (or ISL encapsulation) even though both switches only run VLAN1? Or does 802.1Q tagging/ISL encapsulation only happen when a switch realizes it has more than one VLAN configured on it?
I am trying to understand whether a switch always tags/encapsulates a frame before sending it over a trunk (even when the sending & receiving switch has only VLAN1 configured)....thanks!

ColbyG

If the trunk is dot1q and VLAN 1 is the native VLAN it won't be tagged.

up2thetime

ColbyNA wrote: »

If the trunk is dot1q and VLAN 1 is the native VLAN it won't be tagged.

Thank you!

Just another question.. I thought switches by default try to negotiate a trunk..

I just connected 2 switches in packet tracer and here is the output:
Switch#show interfaces switchport
Name: Fa0/1
Switchport: Enabled
Administrative Mode: dynamic auto
Operational Mode: static access
Administrative Trunking Encapsulation: dot1q
Operational Trunking Encapsulation: native
Negotiation of Trunking: On
Access Mode VLAN: 1 (default)
Trunking Native Mode VLAN: 1 (default)
Voice VLAN: none
Administrative private-vlan host-association: none
...
...
Why does admin mode say dynamic auto.. but operational mode says static access... i never did any configuration on these switches (except for vtp domain name and vlan creation on switch 1).. just connected them.. i expected Operational Mode to say Trunk (especially since Admin Mode says Dynamic Auto)..

I was wondering why Switch2 was not getting the VLAN database i configured on switch one (vtp domain was configured on both).. i guess this is because the connection between the switch is functioning as Static Access.. i just dont understand why this is...Negotiating of Trunking is ON as well... why didnt they auto negotiate a trunk?

mikej412

ISL Trunks will always tag traffic

Trunking depends on the switches.

2924-XL-EN switches don't support DTP, so you have manually tell a port to trunk.

2950 switches will attempt to trunk by default -- but only support dot1q. I assume the 2960 is the same.

3550 switches default to dynamic desirable -- so they will trunk with other 3550 switches -- and 3560 switches and 3750 switches which default to dynamic auto.

3560 (and 3750) default to dynamic auto -- and 2 dynamic auto switches will NOT negotiate a trunk. You'd have to change the default configuration on at least one switch to get it to trunk.

up2thetime

mikej412 wrote: »

ISL Trunks will always tag traffic

Trunking depends on the switches.

2924-XL-EN switches don't support DTP, so you have manually tell a port to trunk.

2950 switches will attempt to trunk by default -- but only support dot1q. I assume the 2960 is the same.

3550 switches default to dynamic desirable -- so they will trunk with other 3550 switches -- and 3560 switches and 3750 switches which default to dynamic auto.

3560 (and 3750) default to dynamic auto -- and 2 dynamic auto switches will NOT negotiate a trunk. You'd have to change the default configuration on at least one switch to get it to trunk.

thanks a lot mike .. i configured the ports with switchport mode trunk.. and the vlan database passed through!!

this also leads me to 2 other questions i cannot seem to figure out...

(1)
trunks pass all vlan traffic...so...
if switch1 has two PCs in vlan 300.. PC1 (on fa0/1) has mac address aaaa.aaaa.aaaa and PC2 (on fa0/2) has mac bbbb.bbbb.bbbb
switch1 also connects to switch2 on a trunk link
when PC1 sends to PC2.. does the switch forward the frame out fa0/2? (assume the CAM table has all relevant mac addresses listed)
or does it send it out fa0/2 AND the trunk (since trunks pass all vlan traffic)
assume vtp pruning is not configured...
it wouldnt make sends for it to send it across the trunk...since it knows PC2 is located of its own fa0/2



(2)
when a switch receives frames on a trunk link, does it examine the source mac address and add to its CAM table?
example... switch1 and switch2 connect via a trunk
PC1 on switch1 in vlan 500 with mac address aaaa.aaaa.aaaa sends data to a PC2 in vlan 500 on switch 2..
switch 2 will then learn that MAC aaaa.aaaa.aaaa is located off the trunk... so when anyone on switch 2 has a frame for aaaa.aaaa.aaaa the switch sends it out the trunk
here is what i dont understand... is switch2 sending the frame destined to aaaa.aaaa.aaaaa because (a) trunks pass all vlan traffic? or because (b) the switch knows aaaa.aaaa.aaaa is located off its trunk

mella060

Remember that when both sides of the link are set to dynamic auto then no trunk will form because neither end will be sending DTP packets (basically used to attempt for form a trunk link). The dynamic auto interface becomes a trunk interface if its neighboring interface is set to trunk or desirable mode.

Dynamic desirable means that the interface desires or actively attempts to convert the link to a trunk link. The interface becomes a trunk link if the neighboring interface is set to trunk, desirable or auto mode.

Most newer switch models now default to dynamic auto to stop unnecessary trunk links from forming.

Honestly, its pretty rare that you'd continue to use the ISL encapsulation method. Cisco is moving away from ISL. That's why you find that switch models like the 2950 and newer ones don't even support it.

billscott92787

up2thetime wrote: »

Thank you!

Just another question.. I thought switches by default try to negotiate a trunk..

I just connected 2 switches in packet tracer and here is the output:
Switch#show interfaces switchport
Name: Fa0/1
Switchport: Enabled
Administrative Mode: dynamic auto
Operational Mode: static access
Administrative Trunking Encapsulation: dot1q
Operational Trunking Encapsulation: native
Negotiation of Trunking: On
Access Mode VLAN: 1 (default)
Trunking Native Mode VLAN: 1 (default)
Voice VLAN: none
Administrative private-vlan host-association: none
...
...
Why does admin mode say dynamic auto.. but operational mode says static access... i never did any configuration on these switches (except for vtp domain name and vlan creation on switch 1).. just connected them.. i expected Operational Mode to say Trunk (especially since Admin Mode says Dynamic Auto)..

I was wondering why Switch2 was not getting the VLAN database i configured on switch one (vtp domain was configured on both).. i guess this is because the connection between the switch is functioning as Static Access.. i just dont understand why this is...Negotiating of Trunking is ON as well... why didnt they auto negotiate a trunk?

Your config showed:

Administrative Mode: dynamic auto
Operational Mode: static access

You said you didn't configure anything except VTP, that's the reason they aren't sharing the VLAN information. By default the switch you are using defaults to dynamic auto, remember if "BOTH" ends of the link are set to auto, then a trunk is not going to form. It will just sit there. It shows static access because that's what it defaults to, it is an access port basically right now, except trunking can occur if the other side of the link initiates the trunking, but it isn't going to since both sides are dynamic auto, if one was set to trunking:

switchport mode trunk
or
switchport mode dynamic desirable

Then the trunk would form and they would share VLAN informatin. As long as you have them in the same VTP domain and password (if one is set).











Here are your other questions:


(1)
trunks pass all vlan traffic...so...
if switch1 has two PCs in vlan 300.. PC1 (on fa0/1) has mac address aaaa.aaaa.aaaa and PC2 (on fa0/2) has mac bbbb.bbbb.bbbb
switch1 also connects to switch2 on a trunk link
when PC1 sends to PC2.. does the switch forward the frame out fa0/2? (assume the CAM table has all relevant mac addresses listed)
or does it send it out fa0/2 AND the trunk (since trunks pass all vlan traffic)
assume vtp pruning is not configured...
it wouldnt make sends for it to send it across the trunk...since it knows PC2 is located of its own fa0/2


ANSWER:
Remember if the address is in the MAC table, it is only going to forward it out the interface on which that MAC address is associated with. So, since the switch knows the MAC of PC2, it will forward it out the fa0/2 interface. If you just configured the switches and pinged for the first time, then it will send it out fa0/2 and the trunk port. But, when the switch over the trunk gets it, if there's no MAC associated with it (which in this case there isn't) it will drop the frame. Remember if you have packet tracer you can lab this out, you can even check it out in the simulation mode and watch the frames pass and how they react in a particular situation like this. Just keep that in mind you can help yourself out by labbing it out. It seems you grasp it more. At least that's how I am.


(2)
when a switch receives frames on a trunk link, does it examine the source mac address and add to its CAM table?
example... switch1 and switch2 connect via a trunk
PC1 on switch1 in vlan 500 with mac address aaaa.aaaa.aaaa sends data to a PC2 in vlan 500 on switch 2..
switch 2 will then learn that MAC aaaa.aaaa.aaaa is located off the trunk... so when anyone on switch 2 has a frame for aaaa.aaaa.aaaa the switch sends it out the trunk
here is what i dont understand... is switch2 sending the frame destined to aaaa.aaaa.aaaaa because (a) trunks pass all vlan traffic? or because (b) the switch knows aaaa.aaaa.aaaa is located off its trunk

ANSWER:

First thing, REMEMBER ALWAYS, ALWAYS, as far as I am aware, a switch examines each frame and compares it to it's MAC address table, based on "SOURCE MAC ADDRESS." If that address is not in the MAC address table, then it adds it and it's associated interface on which the frame was received." Next, it processes where to send the frame, it will look for the destination MAC address, if it sees it, it forwards it out that interface. If not, it is broadcasted out all interfaces except the one on which it was received.

Correct me if I am wrong everyone, But, it is passing it out that interface because the switch knows that aaaa.aaaa.aaaa.aaaa is located off of the trunk interface. It already has the address at this point (based on his description) in the MAC table. If the VLAN wasn't allowed to pass traffic for that particular VLAN over that interface, it wouldn't know the PC's mac address to begin with.

up2thetime

billscott92787 wrote: »

Your config showed:

Administrative Mode: dynamic auto
Operational Mode: static access

You said you didn't configure anything except VTP, that's the reason they aren't sharing the VLAN information. By default the switch you are using defaults to dynamic auto, remember if "BOTH" ends of the link are set to auto, then a trunk is not going to form. It will just sit there. It shows static access because that's what it defaults to, it is an access port basically right now, except trunking can occur if the other side of the link initiates the trunking, but it isn't going to since both sides are dynamic auto, if one was set to trunking:

switchport mode trunk
or
switchport mode dynamic desirable

Then the trunk would form and they would share VLAN informatin. As long as you have them in the same VTP domain and password (if one is set).











Here are your other questions:


(1)
trunks pass all vlan traffic...so...
if switch1 has two PCs in vlan 300.. PC1 (on fa0/1) has mac address aaaa.aaaa.aaaa and PC2 (on fa0/2) has mac bbbb.bbbb.bbbb
switch1 also connects to switch2 on a trunk link
when PC1 sends to PC2.. does the switch forward the frame out fa0/2? (assume the CAM table has all relevant mac addresses listed)
or does it send it out fa0/2 AND the trunk (since trunks pass all vlan traffic)
assume vtp pruning is not configured...
it wouldnt make sends for it to send it across the trunk...since it knows PC2 is located of its own fa0/2


ANSWER:
Remember if the address is in the MAC table, it is only going to forward it out the interface on which that MAC address is associated with. So, since the switch knows the MAC of PC2, it will forward it out the fa0/2 interface. If you just configured the switches and pinged for the first time, then it will send it out fa0/2 and the trunk port. But, when the switch over the trunk gets it, if there's no MAC associated with it (which in this case there isn't) it will drop the frame. Remember if you have packet tracer you can lab this out, you can even check it out in the simulation mode and watch the frames pass and how they react in a particular situation like this. Just keep that in mind you can help yourself out by labbing it out. It seems you grasp it more. At least that's how I am.


(2)
when a switch receives frames on a trunk link, does it examine the source mac address and add to its CAM table?
example... switch1 and switch2 connect via a trunk
PC1 on switch1 in vlan 500 with mac address aaaa.aaaa.aaaa sends data to a PC2 in vlan 500 on switch 2..
switch 2 will then learn that MAC aaaa.aaaa.aaaa is located off the trunk... so when anyone on switch 2 has a frame for aaaa.aaaa.aaaa the switch sends it out the trunk
here is what i dont understand... is switch2 sending the frame destined to aaaa.aaaa.aaaaa because (a) trunks pass all vlan traffic? or because (b) the switch knows aaaa.aaaa.aaaa is located off its trunk

ANSWER:

First thing, REMEMBER ALWAYS, ALWAYS, as far as I am aware, a switch examines each frame and compares it to it's MAC address table, based on "SOURCE MAC ADDRESS." If that address is not in the MAC address table, then it adds it and it's associated interface on which the frame was received." Next, it processes where to send the frame, it will look for the destination MAC address, if it sees it, it forwards it out that interface. If not, it is broadcasted out all interfaces except the one on which it was received.

Correct me if I am wrong everyone, But, it is passing it out that interface because the switch knows that aaaa.aaaa.aaaa.aaaa is located off of the trunk interface. It already has the address at this point (based on his description) in the MAC table. If the VLAN wasn't allowed to pass traffic for that particular VLAN over that interface, it wouldn't know the PC's mac address to begin with.

Correct you are billscott!

Traced the packets and examined the CAM tables!

Only the ARPs went over the trunk (and then only when the ICMP had to be sent over the trunk)

Thank you!!

billscott92787

No prob. Keep that in mind for future questions. You can definitely learn a lot of how things work by labbing them out and then watching how it works in packet tracer using the simulation side. It helps you really build the "theory" behind it. Since you can't really get that with the networking equipment itself. I mean you can do debugs and things like that but, this gives you a visual example.