Add dp_channel_tunnel_key attribute to HA set

[marian-pritsak]

According to the 4.2.1.2. Forwarding packet to remote DPU section 1, the tunneled packets MUST have an VxLan encap as the outer packet, with a specified NPU tunnel VNI number as identifier.

This PR adds the HA set attribute to specify the VNI of the tunnel.

[mssonicbld]

/azp run

[azure-pipelines]

Commenter does not have sufficient privileges for PR 679 in repo sonic-net/DASH

[prsunny]

@marian-pritsak , 4.2.1.2 section is for NPU tunnels. Are you referring to DPU originated tunnels or from NPU?
@r12f , @zjswhhh for review

[r12f]

I believe this is the tunnel used for inline sync. In the HLD it is designed to be a UDP encapped instead of VxLan encapped, hence there is no requirement and attribute being added for this field. Is there any issue being found with the original design?

Also @prsunny , are you aware of any issues with the current design as well?

[KrisNey-MSFT]

Hi @mukeshmv and @prsunny - any thoughts on this one?

[KrisNey-MSFT]

@mukeshmv and @prsunny - any thoughts here?

[mukeshmv]

@mukeshmv and @prsunny - any thoughts here?

@KrisNey-MSFT, @r12f seems like there is some confusion.
The smartswitch HA HLD Section 4.2.12 referenced in this PR talks about NPU-NPU tunnel.
But this PR has added an attribute in HA Set for dp-channel which is the DPU-DPU tunnel.
If the intention is to specify the NPU-NPU tunnel VNI then I am not sure if the HA-set object is the right place.
Can we reuse any existing SAI VXLAN tunnel key attribute for the NPU.

[KrisNey-MSFT]

@mukeshmv and @prsunny - any thoughts here?

@KrisNey-MSFT, @r12f seems like there is some confusion. The smartswitch HA HLD Section 4.2.12 referenced in this PR talks about NPU-NPU tunnel. But this PR has added an attribute in HA Set for dp-channel which is the DPU-DPU tunnel. If the intention is to specify the NPU-NPU tunnel VNI then I am not sure if the HA-set object is the right place. Can we reuse any existing SAI VXLAN tunnel key attribute for the NPU.

@marian-pritsak for viz, as he is the author...

[KrisNey-MSFT]

hi @marian-pritsak and @oleksandrivantsiv - do we still need this PR after looking at @mukeshmv 's comments? TY, Kristina