TrunkedL2 and Routed3 Link in WAN

hendychow

Hello All,

I have a question about VLAN, LAN and WAN.
Does WAN always use Layer 3 (Routed link) ?
Does VLAN and LAN always use Layer 2 (Trunked link) ?

The reason I am asking the basic thing is because in my company where I work,
there are both routed and trunked link between the core (multilayer switch) in WAN connection (which use Gigabit ethernet connection / leased line)

That means possibly we send VLAN across WAN (I am not sure)

Now, is it possible if we just use Layer 2 in WAN connection (using multilayer core switch) ?
If yes, do we still need to use Routing Protocol ?

Thank You,

Hendy

malcybood

hendychow wrote: »

Hello All,

I have a question about VLAN, LAN and WAN.
Does WAN always use Layer 3 (Routed link) ?
Does VLAN and LAN always use Layer 2 (Trunked link) ?

The reason I am asking the basic thing is because in my company where I work,
there are both routed and trunked link between the core (multilayer switch) in WAN connection (which use Gigabit ethernet connection / leased line)

That means possibly we send VLAN across WAN (I am not sure)

Now, is it possible if we just use Layer 2 in WAN connection (using multilayer core switch) ?
If yes, do we still need to use Routing Protocol ?

Thank You,

Hendy

The easiest way to get this in your head is to think that a layer 3 link requires a layer 2 technology to operate and for a packet to get to it's destination over a WAN, requires a routing protocol or static routing.

You traverse a WAN, ultimately to get to another LAN so picture it like this

LAN "A" (layer2) - WAN "A" (Layer2/3) - Service Provider "cloud" - WAN "B" (Layer 2/3) - LAN "B"

this could also be described as

LAN Switch "A" - WAN Router "A" - Service Provider "cloud" - WAN Router "B" - LAN Switch "B"

So when you leave LAN "A" the packet is going into "the unknown" until the routing protocol on WAN "A" router, directs the packet over the cloud to the other side of the WAN via IP subnet & destination IP address.

When the packet gets to the remote WAN the router then says, "oh you have a destination address of LAN B, I'll send you there".

In regard to VLAN trunking, this would be configured between the LAN switch and the router if you required traffic from 2 or more VLAN's to travel over the WAN through the router. This is achieved by dot1q trunking and there's lots of online resources which explains how dot1q and inter VLAN routing works.

Hopefully this gives a simple analogy for you as to how it works and I've not confused you more!

hendychow

Hello malcybood,

Thank you for your response, but I think you missed my point.
Here is the scenario:

MLS = Multilayer Switch 6500
S = Server

S ---- MLS A ---- MLS B

---

WAN

---

MLS C

|

---

Data Center

---

|

S, MLS A and MLS B are located in the Data Center (same location).
MLS B does the Layer 3 routing to WAN

MLS A and MLS B has a trunked link (Layer 2), let say
interface gigethernet 2/3
switchport trunk allowed vlan 123
switchport mode trunk
and also

routed link (Layer 3) through Switch Virtual Interface, let say
interface Vlan123
ip address x.x.x.x x.x.x.x
ip ospf network point-to-point
end

Since I am only a support guy, so I don't really know why the network architect did this (I never get a chance to ask)

My question is:
- Why are there Trunked (Layer 2) and Routed (Layer 3) on the same link ?

- Is it possible to use Layer 2 (Trunked link) across WAN ? (I am not talking about MPLS here)

Thank You so much,

Hendy

malcybood wrote: »

The easiest way to get this in your head is to think that a layer 3 link requires a layer 2 technology to operate and for a packet to get to it's destination over a WAN, requires a routing protocol or static routing.

You traverse a WAN, ultimately to get to another LAN so picture it like this

LAN "A" (layer2) - WAN "A" (Layer2/3) - Service Provider "cloud" - WAN "B" (Layer 2/3) - LAN "B"

this could also be described as

LAN Switch "A" - WAN Router "A" - Service Provider "cloud" - WAN Router "B" - LAN Switch "B"

So when you leave LAN "A" the packet is going into "the unknown" until the routing protocol on WAN "A" router, directs the packet over the cloud to the other side of the WAN via IP subnet & destination IP address.

When the packet gets to the remote WAN the router then says, "oh you have a destination address of LAN B, I'll send you there".

In regard to VLAN trunking, this would be configured between the LAN switch and the router if you required traffic from 2 or more VLAN's to travel over the WAN through the router. This is achieved by dot1q trunking and there's lots of online resources which explains how dot1q and inter VLAN routing works.

Hopefully this gives a simple analogy for you as to how it works and I've not confused you more!

rossonieri#1

hi hendy,

nice questions

btw - malcybood has a good point.

So when you leave LAN "A" the packet is going into "the unknown" until the routing protocol on WAN "A" router, directs the packet over the cloud to the other side of the WAN via IP subnet & destination IP address.

When the packet gets to the remote WAN the router then says, "oh you have a destination address of LAN B, I'll send you there".

My question is:
- Why are there Trunked (Layer 2) and Routed (Layer 3) on the same link ?

i hope you dont mind to answer your own questions by doing some thinking a little bit to review our past CCNA study - shall we :
why do we need a switch or a router exactly?
why cant we just use that L2 switching or use that trunk line?

- Is it possible to use Layer 2 (Trunked link) across WAN ? (I am not talking about MPLS here)

ok, no - we dont talk MPLS yet.
lets review our basic - this is the best part : what are those exactly WAN or LAN or MAN or GAN? are those a prerequisite to do something?

can we just put our 1000miles fiber optic between 2 sites we call it LAN? or should we call it a WAN?

and how about that frame-relay network for example? why we need IPs?

i hope that you catch the idea.

Thank You so much,

you are very welcome

hendychow

Hello rossonieri#1,

I guess I am confused
I am not smart enough to comprehend your explanation, but I really appreciate your respond.

If you could, please explain it straightforward

I understood malcybood's explanation, but It's just a basic concept and It did not explain anything what had happened on the topology

My question is not a make-up topology, this involves a "real world" scenario.
We have 30 cores routers and about 2000 switch and 500 servers.. It is a very complex scenario.... (Juniper, Cisco, Load balancer, fwsm, ASA, etc)
Unfortunately, I only met the "architect" (4 digit CCIE) once... and he did not have time to explain it to a CCNA guy like me

I guess most the people in this forum only cares about CCNP

If we can talk about how to utilize CCNP in real-world, it would be more interesting

So.. anybody can explain why Layer 2 and Layer 3 are in the same link in LAN ?
Are both can be used ?

Thanks,

Hendy

rossonieri#1 wrote: »

hi hendy,

nice questions

btw - malcybood has a good point.




i hope you dont mind to answer your own questions by doing some thinking a little bit to review our past CCNA study - shall we :
why do we need a switch or a router exactly?
why cant we just use that L2 switching or use that trunk line?



ok, no - we dont talk MPLS yet.
lets review our basic - this is the best part : what are those exactly WAN or LAN or MAN or GAN? are those a prerequisite to do something?

can we just put our 1000miles fiber optic between 2 sites we call it LAN? or should we call it a WAN?

and how about that frame-relay network for example? why we need IPs?

i hope that you catch the idea.



you are very welcome

rossonieri#1

hi hendy,

i feel sorry that you still cant catch the idea.

I understood malcybood's explanation, but It's just a basic concept and It did not explain anything what had happened on the topology

both malcy and me - perhaps - trying to explain to you questions mainly on "why there are L3 part on L2 link". including examples.

My question is not a make-up topology, this involves a "real world" scenario.

define real-world

well, it wouldnt go nowhere if we keep saying which "real world" and which arent - would we? i pretty much sure that you guys (especially in US backbone side) are having much more equipment compares to my side in 3rd world countries but that is not how it goes, the thing is do you know what you have/do.

Unfortunately, I only met the "architect" (4 digit CCIE) once... and he did not have time to explain it to a CCNA guy like me

your questions did not relate to IE digit number - believe me, in my side there are bunch of 4 digit IE too ) so, next time you meet your 4 digit IE - go bring him some lunch, go chat around about things other than his work - just go slow - be nice with him, and you'll go deeper after that )

or you can ask our TE - IEs like dt, or mrD and others NP/IP level friends )

I guess most the people in this forum only cares about CCNP
If we can talk about how to utilize CCNP in real-world, it would be more interesting

come on ) this is not a true networker attitude. just do what i've told you above )

So.. anybody can explain why Layer 2 and Layer 3 are in the same link in LAN ?
Are both can be used ?

i guess i'm done - i'm sorry - i couldnt explain any better
so, wish you luck in reviewing your questions

no offense,

cheers

rakem

hendychow wrote: »

Hello malcybood,

Thank you for your response, but I think you missed my point.
Here is the scenario:

MLS = Multilayer Switch 6500
S = Server

S ---- MLS A ---- MLS B

---

WAN

---

MLS C

|

---

Data Center

---

|

S, MLS A and MLS B are located in the Data Center (same location).
MLS B does the Layer 3 routing to WAN

MLS A and MLS B has a trunked link (Layer 2), let say
interface gigethernet 2/3
switchport trunk allowed vlan 123
switchport mode trunk
and also

routed link (Layer 3) through Switch Virtual Interface, let say
interface Vlan123
ip address x.x.x.x x.x.x.x
ip ospf network point-to-point
end

Since I am only a support guy, so I don't really know why the network architect did this (I never get a chance to ask)

My question is:
- Why are there Trunked (Layer 2) and Routed (Layer 3) on the same link ?

- Is it possible to use Layer 2 (Trunked link) across WAN ? (I am not talking about MPLS here)

Thank You so much,

Hendy

Ok my answer to this is simple. Its a layer 3 link to stop broadcasts getting sent across the WAN.

interface vlan123 is the default gateway for your network and its is using a layer three link to segment the network. There is really no reason to have a layer 2 link over a WAN. Even if it was a fibre connection to a site that could be seen as a layer 2 link i would still route it.

networker050184

Google metro ethernet.