VLAN trunking question

Hello,

Do I need to use trunk between SW5 & SW4, if VLAN5 is laid separately as shown in attached picture ? That is, if it is used more than one VLAN in İntranet, does it mean I have to use trunk mode between ALL of switch connections ?

And second question. Let's suppose I placed my DHCP server in a separate VLAN which is Native. Shall I use subinterface encapsulation dot1q 1 (native vlan) or It's sufficient to use ip dhcp pool name & ip helper-address commands without tagging it ??

Thanks in advance

forumq.JPG

Find more posts tagged with

Comments

Jon_Cisco

Start by thinking about what a vlan is doing. It separates traffic.
Now figure out where that traffic needs to go.

Does vlan 1-4 need to reach 5? Does vlan 5 need to reach 1-4? If so how are they changing networks?

So by reading your question I don't think you should be asking if trunks need to exist between all switches. I think you need to consider what the trunk actually does on the switch. Then figure out where it's needed.

Welcome to the forums!

Deathmage

Me personally, I'd remove the daisy-chainned trunk(s) on the 2960's and move the 2950 to one of the vlans and move the now freed up 2960 as the core L3 routing switch to the 1841, then connect the DHCP server to the 2960 'routing switch' (that DHCP sever would need to be on a vlan too, an IP address is kind of helpful for the 'ip helper-address' command) then have the 2960 connected to the 1841.

My reasoning is this, even though the 2960 is a L3 having all those connection daisy-chainned into themselves can be a interesting routing table, so just make the access-layer 2960's L2's and make one of then at the distribution layer a L3. To me that would flow allot better than your above config, the above config would work, just from a learning standpoint changing the topology would be easier to understand.

So to recap, the 1841 would be the core level, the 2960 with L3 function turned on would be the Distribution layer (with DHCP on this switch on a vlan for IP Address-helper placement) and then the lower level of 2960's and the 2950 at L2 would be the access layer.

Now something to munch on if this is just a test lab, you can do DHCP on the 1841....

==============================
Now on the access layer, the basic trunk config would be something like this on each L2 switch:

Interface fastethernet 0/1
switchport mode trunk
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 1 (note 1)
switchport trunk native vlan 888

vlan 888
description "disabled ports and native vlan"
no ip address
shutdown

===============================

note 1: (each L2 switch would have a different number - unless you want all of the L2 switches to talk to each other with inter-vlan routing, then the 'allowed' would encompass them all) - turning on 'ip routing' on the L3 2960 would be needed for inter-vlan routing if this is the desired result.

Phileeeeeeep651

Not to get off topic from the OP's question but I thought the 2960's were L2 switches?

Deathmage

Phileeeeeeep651 wrote: »

Not to get off topic from the OP's question but I thought the 2960's were L2 switches?

There L2 switches with static routes so I suppose they could be a L2+ switch since they support limited routing, as of Cisco iOS 12.2. Typically a switch that can do routing is considered a L3 switch, but I'll leave the previous sentence for those that are curious to make there own determination.

Here is an article on the topic over @ Cisco:
http://www.cisco.com/en/US/docs/switches/lan/catalyst2960/software/release/12.2_55_se/configuration/guide/swipstatrout.html.

Phileeeeeeep651

Thanks for the info and the link! You learn something new every day!

showint

Thank everyone for your replies. I could find an answer to my question.

How can I choose the best answer here?