Lab Objective
Configure LACP fast rate on an EtherChannel bundle's member interfaces, verify the reduced LACPDU interval takes effect, and compare the time to detect a failed member link against the default slow rate behavior.
Lab Purpose
LACP, discussed earlier in this series regarding EtherChannel negotiation, periodically exchanges control packets to confirm each member link remains healthy. The default interval is relatively slow, meaning a failed member link can take longer than necessary to be detected and removed from the bundle — LACP fast rate reduces this detection time significantly.
Lab Topology
Switch1 ---- Po1 (Gi1/0/1, Gi1/0/2) ---- Switch2
LACP already configured and bundle active,
as covered in an earlier labTask 1: Verify the Default LACP Rate
Confirm the current LACPDU interval on the bundle's member interfaces.
Task 2: Configure LACP Fast Rate
Enable LACP fast rate on both member interfaces on Switch1.
Task 3: Verify the New Rate Takes Effect
Confirm the LACPDU interval has changed accordingly.
Task 4: Simulate a Member Link Failure Under the Default Rate
Before applying fast rate on Switch2's side, simulate a link failure and time how long detection takes under the default slow rate.
Task 5: Apply Fast Rate on Switch2 and Repeat the Failure Test
Configure fast rate on Switch2 as well, then repeat the failure simulation and compare detection time.
Solution and Verification
Switch1# show lacp internal
Flags: S - Device is requesting Slow LACPDUs
Port Flags State LACP port Priority
Gi1/0/1 SA bndl 32768
Gi1/0/2 SA bndl 32768
-- "S" confirms Slow LACPDUs are currently
-- being requested -- the default interval
-- of 30 seconds between LACP hellosSwitch1(config)# interface range gigabitethernet1/0/1-2
Switch1(config-if-range)# lacp rate fastSwitch1# show lacp internal
Flags: F - Device is requesting Fast LACPDUs
Port Flags State LACP port Priority
Gi1/0/1 FA bndl 32768
Gi1/0/2 FA bndl 32768
-- "F" confirms fast rate is now active,
-- reducing the LACPDU interval from 30
-- seconds down to 1 second-- With Switch2 still at the default slow
-- rate, simulating Gi1/0/1's link failure
-- (e.g., unplugging or shutting the port):
-- Detection and removal from the bundle
-- took up to approximately 90 seconds
-- (roughly three missed slow-rate intervals
-- before the peer is considered down),
-- since Switch2's side was still operating
-- on the default 30-second intervalSwitch2(config)# interface range gigabitethernet1/0/1-2
Switch2(config-if-range)# lacp rate fast-- Repeating the same link failure
-- simulation with fast rate now active
-- on both sides:
-- Detection and removal from the bundle
-- occurred within roughly 3 seconds
-- (three missed 1-second fast-rate
-- intervals), a dramatic improvement over
-- the default rate's detection timeKey Takeaway
LACP fast rate is configured independently on each side of a link, but both sides typically need it enabled to see the full benefit, since detection timing depends on the interval the LOCAL switch uses to expect hellos from its peer — configuring it only on one switch still leaves that switch waiting on the slower default timing from its neighbor's perspective, exactly as demonstrated by the intermediate test in this lab before Switch2 was updated.