switchport trunk encapsulation on L3 switches

[chirschey]

                Why is 'switchport trunk encapsulation <dot1q or isl> required on L3 switches?  The default trunk encapsuation mode on 'modern' Cisco switches is to 'auto' negotiate, so why doesn't 'auto-negotiate' work when configured from the L3 switch port?  If I configure 'switchport mode trunk' on an L2 switch (capable of only dot1q) and don't configure the adjacent L3 port, the trunk is auto-negotiated.  However, if I configure 'switchport mode trunk' on the L3 port first, it gives the error we've all witnessed: Command rejected: An interface whose trunk encapsulation is "Auto" can not be configured to "trunk" mode. Interestingly, if I configure, 'switchport mode dynamic desirable' on the L3 port, the interface does indeed negotiate the trunk encapsulation and establish the trunk.  According to Cisco documentation, the 'switchport mode trunk' command is also supposed to negotiate the trunking status and encapsulation--so why doesn't this command work the same as 'switchport mode dynamic desirable?'

[John Blakley]

Chris,

On later switches, you dont' have the option of setting encapsulation at all because Cisco is moving away from ISL encapsulation. The dynamic desirable mode, as you know, dynamically forms a trunk if the neighbor on the other end can agreen to also for a trunk via DTP frames. This, by default, usually ends up in ISL encapsulation if both switches support it. The trunk command is static, and because it's static, you also need to tell the switch what you are expecting to support - whether that be ISL or dot1q. On later switches, you can set to trunk without setting encapsulation because the switch will only support dot1q.

HTH,
John

*** Please rate all useful posts ***

[chirschey]

John,

You're absolutely correct.  My hope is that Cisco will change its definition for 'switchport mode trunk.'

This is from their documentation:

switchport mode dynamic desirable

Makes the interface actively attempt to convert the link to a trunk link. The interface becomes a trunk interface if the neighboring interface is set to trunk, desirable, or auto mode.

switchport mode trunk

Puts the interface into permanent trunking mode and negotiates to convert the neighboring link into a trunk link. The interface becomes a trunk interface even if the neighboring interface is not a trunk interface.

switchport nonegotiate

Prevents the interface from generating DTP frames. You can use this command only when the interface switchport mode is access or trunk. You must manually configure the neighboring interface as a trunk interface to establish a trunk link.

I've highlighted negotiates to point out that DTP frames are still sent to the neighboring device to negotiate the trunking status. Therefore, why doesn't it also negotiate the encapsulation type when desiring to trunk? My point being, if it's going to trunk unconditionally and not negotiate the trunking protocol, and since you'd have to have an ISL-only switch (non-extant), Cisco should simply get rid of ISL on their switches or have the 'negotiation' process or (unconditional state) select dot1Q as the trunking protocol.