This is to test the functionality of decapsulation of GUEv1 to IPv4 or IPv6 payload and ensures that only the outer header (IPv4 | UDP) of GUEv1 encapsulated packets are used for hashing on decapsulating nodes.

GUEv1 decapsulation matching destination-address-prefix-set and TTL and DSCP behavior test is documented in #4178.

graph LR; 
subgraph DUT [DUT]
    B1[Port1]
    B2[Port2]
    B3[Port3]
    B4[Port4]
    B5[Port5]
    B6[Port6]
    B7[Port7]
end
subgraph ATE2 [ATE2]
    N1[Port1]
end
subgraph ATE3 [ATE3]
    N21[Port2]
    N22[Port3]
end
subgraph ATE4 [ATE4]
    N31[Port4]
    N32[Port5]
end
subgraph ATE5 [ATE5]
    N4[Port6]
end
A1[ATE1:Port1] <--EBGP--> B1; 
B2 <--IBGP--> N1; 
B3 <-- [LAG1] IBGP --> N21;
B4 <-- [LAG1] IBGP --> N22;
B5 <-- [LAG2] EBGP --> N31; 
B6 <-- [LAG2] EBGP --> N32;
B7 <-- EBGP --> N4; 

1. The DUT's loopback interface must be passive for IS-IS.
2. Configure IPv4 and IPv6 EBGP[ASN200:ASN100] between ATE:Port1 <> DUT:Port1
3. Configure IPv4 and IPv6 IBGP[ASN100] between DUT <> ATE2
4. Configure IPv4 and IPv6 IBGP[ASN100] between DUT <> ATE3
5. Configure IPv4 and IPv6 EBGP[ASN100:ASN200] between DUT <> ATE4
6. Configure IPv4 and IPv6 EBGP[ASN100:ASN200] between DUT <> ATE5
7. Configure DUT as decapsulation node for IP|UDP (GUE v1)using "DUT-DECAP-Address" and decap UDP port as 6080
8. Enable BGP multipath for both EBGP and IBGP sessions to enable load balancing traffic across multiple paths/links
9. Enable BGP multihop for BGP(IBGP/EBGP) sessions on LAG interfaces
10. DUT has multiple paths to Host2 via multiple nodes, ATE2 & ATE3
11. DUT can reach Host3 via a lag interface towards ATE3
12. DUT has multiple paths to Host4 via multiple nodes, ATE4 & ATE5
13. Host1(v4/v6) route is installed and active via ATE1
14. Host2(v4/v6) route is installed and active via ATE2 and ATE3, therefore the traffic for Host2 should be load-balanced across both the nodes
15. Host3v4 route is installed and active via ATE3, therefore the traffic for Host3 should be load-balanced across the lag members
16. Host4(v4/v6) route is installed and active via ATE4 and ATE5, therefore the traffic for Host4 should be load-balanced across both the nodes

1. The ATE's loopback interface must be passive for IS-IS
2. Establish BGP sessions as shown in the DUT configuration section
3. ATE1 hosts Host1v4 and Host1v6
4. ATE2 and ATE3 hosts Host2v4 and Host2v6
5. ATE3 hosts Host3v4
   • Reachability to Host3v4 from ATE3 is via a static MPLS label
6. ATE4 and ATE5 hosts Host4v4 and Host4v6

To simplify this document, Host1, Host2, and Host3 will be referred to as H1, H2, H3 and H4 respectively.

• Traffic towards a destination is spread evenly across nodes and LAGs (if applicable):
  • Tolerance for delta: 5%
• Start the Ixia traffic as specified for test
  • Sent 1000000 packets at the 10% of the line rate.
  • Packets are generated based on different header field entropy which can be defined by the test case
• Repeat each test with the each ATE Flow-type or explicitly mentioned flow-type
• Conduct each of the following test, using a single flow-type with 1024 flows

• Configure the DUT and ATE as stated above
• Initiate a single flow-type and follow the below stated and applicable verification steps
• L4 source port of outer header(GUEv1 encap header) should be randomized for each flow-type that's running
• Repeat the test for all flow-types
• Validations:
• The outer header destination IP of the traffic is the DUT-DECAP-Address and the destination port of the traffic (UDP 6080) matches the configured UDP decap port criteria
• Therefore, DUT will decapsulate the outer header and perform a lookup based on the inner IP address
• The following traffic distribution validations are applicable as per the flow-type that is being tested
  • Flow#1 for H3 should be load-balanced across the lag members via ATE3
  • Flow#2 for H2 should be load-balanced via ATE2 and ATE3
    • Traffic via ATE3 should be load-balanced across the lag members
  • Flow#3 for H2 should be load-balanced via ATE2 and ATE3
    • Traffic via ATE3 should be load-balanced across the lag members
  • Flow#4 for H4 should be load-balanced via ATE4 and ATE5
    • Traffic forwarded towards ATE4 (via LAG2) should be load-balanced across the LAG members
  • Flow#5 for H4 should be load-balanced via ATE4 and ATE5
    • Traffic forwarded towards ATE4 (via LAG2) should be load-balanced across the LAG members
  • Flow#6 for H3 should be load-balanced across the lag members via ATE3
  • Flow#7 for H2 should be load-balanced via ATE2 and ATE3
    • Traffic via ATE3 should be load-balanced across the lag members
  • Flow#8 for H2 should be load-balanced via ATE2 and ATE3
    • Traffic via ATE3 should be load-balanced across the lag members
  • Flow#9 for H4 should be load-balanced via ATE4 and ATE5
    • Traffic forwarded towards ATE4 (via LAG2) should be load-balanced across the LAG members
  • Flow#10 for H4 should be load-balanced via ATE4 and ATE5
    • Traffic forwarded towards ATE4 (via LAG2) should be load-balanced across the LAG members
  • No packet loss should be observed
  • Port traffic counters will be leveraged to verify the loab-balance behavior

• Configure the DUT and ATE as stated above
• Repeat the test for flow-type#11 and flow-type#12 (one at a time)
• Validations:
• The outer header destination UDP port (6080) matches a configured decap port
• The outer header destination IP of the traffic does not matches the locally configured decapsulation address(DUT-DECAP-Address), therefore it does not match the decapsulation criteria for the destination IP
• The DUT will not decapsulate the outer header. Instead, it will perform a lookup based on the outer destination IP address and forward the packets as standard IP traffic
• ATE Port 2 receives 1000000 packets
• No packet loss should be observed

• Configure the DUT and ATE as stated above
• Repeat the test for flow-type#13 and flow-type#14 (one at a time)
• Validations:
• The outer header destination IP of the traffic matches a configured decap IP address
• The outer header destination UDP port (6085) of the traffic does not matches the locally configured decapsulation port(6080), therefore it does not match the decapsulation criteria for the destination port
• The DUT should not decapsulate these packets. Packets should be dropped since no specific policy-forwarding rule exists for unmatched GUE
• The DUT decapsulation counters should not increment for this flow
• The drop counters on DUT will reflect the packets to 1000000
  • These drop packets count will be reflected in DUT control-plane drop counters
• 100% packet loss should be observed on ATE Port 2

• Set up the DUT and ATE as previously specified.
• For each flow type, configure a single source and destination IP address for both the inner and middle IPv4 and IPv6 headers. Specifically:
  • For IPv4 flows that would typically use an H1v4 range, use IPv4prefix1/24.
  • For IPv6 flows that would typically use an H1v6 range, use IPv6prefix1/64.
  • For IPv4 flows that would typically use an H2v4 range, use IPv4prefix11/24.
  • For IPv6 flows that would typically use an H2v6 range, use IPv6prefix11/64.
  • For IPv4 flows that would typically use an H4v4 range, use IPv4prefix31/24.
  • For IPv6 flows that would typically use an H4v6 range, use IPv6prefix31/64.
• Randomize the L4 source port of the inner/middle header (the header immediately following the outer header) for each flow type being tested.
• Test Execution:
  • Initiate each flow type sequentially.
  • Perform the applicable verification steps for each flow.
  • Repeat this process for all flow types, applying the modified field as described above.
• Validation:
  • Expected Outcome: The traffic should not be load-balanced; instead, it would simply forwarded along a single path.
  • Port traffic counters will be leveraged to verify this behavior.
  • Failure Condition: If traffic distribution is observed across multiple paths, the test fails.

• Set up the DUT and ATE as previously specified.
• For each flow type, configure a single destination IP address for both the inner and middle IPv4 and IPv6 headers. Specifically:
  • For IPv4 flows that would typically use an H1v4 range, use IPv4prefix1/24.
  • For IPv6 flows that would typically use an H1v6 range, use IPv6prefix1/64.
  • For IPv4 flows that would typically use an H2v4 range, use IPv4prefix11/24.
  • For IPv6 flows that would typically use an H2v6 range, use IPv6prefix11/64.
  • For IPv4 flows that would typically use an H4v4 range, use IPv4prefix31/24.
  • For IPv6 flows that would typically use an H4v6 range, use IPv6prefix31/64.
• For each flow type, configure a source IP addresses for both the inner and middle IPv4 and IPv6 headers. Specifically:
  • For IPv4 flows that would typically use an H1v4 addresses, continue to use the range IPv4prefix1-10/24.
  • For IPv6 flows that would typically use an H1v6 addresses, continue to use the range IPv6prefix1-10/64.
  • For IPv4 flows that would typically use an ATE1LO1v4 IPv4 address, use the range ATE1LO[1-10] addresses for each applicable flow-type.
• Randomize the L4 source port of the inner/middle header (the header immediately following the outer header) for each flow type being tested.
• Test Execution:
  • Initiate each flow type sequentially.
  • Perform the applicable verification steps for each flow.
  • Repeat this process for all flow types, applying the modified field as described above.
• Validation:
  • Expected Outcome: The traffic should not be load-balanced; instead, it would simply forwarded along a single path.
  • Port traffic counters will be leveraged to verify this behavior.
  • Failure Condition: If traffic distribution is observed across multiple paths, the test fails.

TODO: decap policy to be updated by https://github.com/openconfig/public/pull/1288

{
    "network-instances": {
        "network-instance": {
            "config": {
                "name": "DEFAULT"
            },
            "name": "DEFAULT",
            "policy-forwarding": {
                "policies": {
                    "policy": [
                        {
                            "config": {
                                "policy-id": "decap-policy"
                            },
                            "rules": {
                                "rule": [
                                    {
                                        "sequence-id": 1,
                                        "config": {
                                            "sequence-id": 1
                                        },
                                        "ipv4": {
                                            "config": {
                                                "destination-address-prefix-set": "dst_prefix",
                                                "protocol": "IP_UDP"
                                            }
                                        },
                                        "transport": {
                                            "config": {
                                                "destination-port": 6080
                                            }
                                        }
                                        "action": {
                                            "decapsulate-gue": true
                                        },
                                    },
                                ]
                            }
                        }
                    ]
                }
            }
        }
    }
}

paths:

/network-instances/network-instance/policy-forwarding/policies/policy/config/policy-id:
/network-instances/network-instance/policy-forwarding/policies/policy/rules/rule/config/ipv4/config/destination-address-prefix-set:
/network-instances/network-instance/policy-forwarding/policies/policy/rules/rule/config/ipv4/config/protocol:
/network-instances/network-instance/policy-forwarding/policies/policy/rules/rule/transport/config/destination-port:
/network-instances/network-instance/policy-forwarding/policies/policy/rules/rule/action/decapsulate-gue:

# telemetry
/interfaces/interface/state/counters/out-pkts:
/interfaces/interface/state/counters/out-unicast-pkts:
/components/component/integrated-circuit/pipeline-counters/control-plane-traffic/state/dropped-bytes-aggregate:
/components/component/integrated-circuit/pipeline-counters/control-plane-traffic/state/dropped-aggregate:

rpcs:
  gnmi:
    gNMI.Set:
      union_replace: true
      replace: true
    gNMI.Subscribe:
      on_change: true

• Specify the minimum DUT-type:
  • FFF - fixed form factor