Verify that administratively disabling a single member link of a Link Aggregation Group (LAG) via gNMI causes traffic to redistribute to the remaining active member links with minimal packet loss. The overall LAG must remain up, and BGP/IS-IS sessions running over the LAG must remain stable.

Success Metrics:

• Traffic Redistribution: 100% of traffic is successfully carried by the remaining active links post-drain.
• Redistribution Loss: Packet loss duration during the member link disable is under the target milliseconds (e.g., < 10ms depending on platform expectations).

• TESTBED_DUT_ATE_4LINKS

• Connect ATE port-1 to DUT port-1.
• Connect ATE ports 2 through 4 to DUT ports 2 through 4.
• Configure DUT ports 2-4 and ATE ports 2-4 to be part of a single LAG (e.g., PortChannel1).
• Configure dual-stack IPv4 and IPv6 addressing on DUT port-1 and ATE port-1.
  • IPv4: 192.0.2.1/30 (DUT) and 192.0.2.2/30 (ATE).
  • IPv6: 2001:db8::1/126 (DUT) and 2001:db8::2/126 (ATE).
• Configure dual-stack IPv4 and IPv6 addressing on the LAG interface.
  • IPv4: 198.51.100.1/30 (DUT) and 198.51.100.2/30 (ATE).
  • IPv6: 2001:db8:1::1/126 (DUT) and 2001:db8:1::2/126 (ATE).
• Establish eBGP and IS-IS sessions between ATE and DUT over both the standalone link (port-1) and the LAG interface.
• Advertise 10,000 IPv4 and 10,000 IPv6 routes from the ATE over the LAG BGP sessions.

• Step 1: Generate DUT configuration and push via gNMI Set.
  • Verify that /network-instances/network-instance/protocols/protocol/bgp/neighbors/neighbor/state/session-state is ESTABLISHED for sessions over both the standalone link and the LAG.
  • Verify that /network-instances/network-instance/protocols/protocol/isis/interfaces/interface/levels/level/adjacencies/adjacency/state/adjacency-state is UP for adjacencies over both the standalone link and the LAG.
  • Verify that the LAG /interfaces/interface[name=<PortChannel1>]/state/oper-status is UP and all member links have /interfaces/interface[name=<DUT_port>]/state/oper-status as UP.
• Step 2: Start traffic.
  • Send continuous IPv4 and IPv6 traffic from ATE port-1 destined to the 10,000 IPv4 and 10,000 IPv6 routes advertised over the LAG. Ensure the traffic generator creates large-scale entropy (e.g., 10,000+ unique flows by varying Source/Destination IP and UDP ports) to allow for effective LAG hashing.
  • Verify that the traffic is load-balanced across all three member links (DUT ports 2-4) of the LAG. Each link should carry ~33.3% of the traffic (within a ±5% tolerance).
  • Verify that ATE traffic loss is 0%.
• Step 3: Disable one member link.
  • Use gNMI to set /interfaces/interface[name=<DUT_port-2>]/config/enabled to false.

{
  "interfaces": {
    "interface": [
      {
        "name": "Ethernet2",
        "config": {
          "name": "Ethernet2",
          "enabled": false
        }
      }
    ]
  }
}

• Step 4: Validate state after link drain.
  • Verify that /interfaces/interface[name=<DUT_port-2>]/state/enabled is false.
  • Verify that /interfaces/interface[name=<DUT_port-2>]/state/oper-status transitions to DOWN.
  • Verify that the LAG interface /interfaces/interface[name=<PortChannel1>]/state/oper-status remains UP.
  • Verify that the BGP session-state and IS-IS adjacency-state over the LAG remain stable without flapping.
  • Verify that traffic from ATE port-1 redistributes evenly to the remaining active member links (DUT ports 3 and 4). Each link should carry ~50% of the traffic (within a ±5% tolerance).
  • Calculate loss duration = (Tx frames - Rx frames) / (Tx frame rate). Verify loss duration is under the acceptable threshold.
• Step 5: Re-enable the member link.
  • Use gNMI to set /interfaces/interface[name=<DUT_port-2>]/config/enabled to true.
• Step 6: Validate state after link restore.
  • Verify that /interfaces/interface[name=<DUT_port-2>]/state/enabled is true.
  • Verify that /interfaces/interface[name=<DUT_port-2>]/state/oper-status transitions to UP.
  • Verify that the BGP session-state and IS-IS adjacency-state over the LAG remain stable without flapping.
  • Verify that traffic from ATE port-1 redistributes evenly across all three member links (DUT ports 2-4), with each carrying ~33.3% of the traffic (within a ±5% tolerance).
  • Verify that ATE traffic loss is 0% in steady state.

• Step 1: From the restored state of RT-5.19.1, disable the second member link.
  • Use gNMI to set /interfaces/interface[name=<DUT_port-3>]/config/enabled to false.
• Step 2: Validate state.
  • Verify /interfaces/interface[name=<DUT_port-3>]/state/oper-status transitions to DOWN.
  • Verify LAG /interfaces/interface[name=<PortChannel1>]/state/oper-status remains UP.
  • Verify BGP and IS-IS sessions over the LAG remain stable.
  • Verify that traffic from ATE port-1 redistributes evenly to the remaining active member links (DUT ports 2 and 4) with minimal packet loss. Each link should carry ~50% of the traffic (within a ±5% tolerance).
• Step 3: Re-enable the member link.
  • Use gNMI to set /interfaces/interface[name=<DUT_port-3>]/config/enabled to true.
  • Verify traffic shifts back to load balance across DUT ports 2-4.

• Step 1: Disable all active member links.
  • Use gNMI to set /interfaces/interface/config/enabled to false for DUT ports 2, 3, and 4.
• Step 2: Validate state.
  • Verify the LAG /interfaces/interface[name=<PortChannel1>]/state/oper-status transitions to DOWN.
  • Verify BGP session-state transitions out of ESTABLISHED (e.g., ACTIVE or IDLE) and IS-IS adjacency-state transitions to DOWN.
  • Verify 100% traffic loss.
• Step 3: Re-enable all member links.
  • Use gNMI to set /interfaces/interface/config/enabled to true for DUT ports 2, 3, and 4.
• Step 4: Validate recovery.
  • Verify LAG /interfaces/interface[name=<PortChannel1>]/state/oper-status transitions to UP.
  • Verify BGP session-state returns to ESTABLISHED and IS-IS adjacency-state returns to UP.
  • Verify traffic forwarding resumes at 100% and is load-balanced across all three member links.

paths:
  /interfaces/interface/config/enabled:
  /interfaces/interface/config/name:
  /interfaces/interface/state/enabled:
  /interfaces/interface/state/oper-status:
  /interfaces/interface/aggregation/state/member:
  /network-instances/network-instance/protocols/protocol/bgp/neighbors/neighbor/state/session-state:
  /network-instances/network-instance/protocols/protocol/isis/interfaces/interface/levels/level/adjacencies/adjacency/state/adjacency-state:
rpcs:
  gnmi:
    gNMI.Set:
    gNMI.Subscribe:
      on_change: true

• vRX