For M-LAG networking of CE series switches, how to prevent some ports on a slave switch from entering the Error-Down state upon peer-link failures

15

If M-LAG is used on an Ethernet, VXLAN, or IP network where a CE switch is dual-homed to two PEs, when the peer-link is faulty but the heartbeat status is normal, all interfaces except the management interface, peer-link interface, and stack interface on the slave device enter the Error-Down state. When the peer-link recovers, the M-LAG member interfaces in Error-Down state are restored to the Up state after 2 minutes by default. Other physical interfaces in Error-Down state are restored to the Up state automatically.
In actual networking, however, some uplink interfaces that run routing protocols or are used as DAD heartbeat interfaces do not need to enter the Error-Down state when the peer-link is faulty. You can configure the function on non-M-LAG member interfaces.
# Configure Eth-Trunk 2 on the slave switch to automatically enter the Error-Down state.
system-view
[~HUAWEI] interface eth-trunk 2
[*HUAWEI-Eth-Trunk2] m-lag unpaired-port suspend
[*HUAWEI-Eth-Trunk2] commit
- When the peer-link is faulty but the heartbeat status is normal, M-LAG member interfaces and only the interfaces having the m-lag unpaired-port suspend command configured enter the Error-Down state.
- M-LAG member interfaces do not support the m-lag unpaired-port suspend command, so you are advised to configure them on both master and slave switches.

Other related questions:
CE series switch Error-Down mechanism
For more information about the CloudEngine Series Switches Error-Down Mechanism, click CloudEngine Series Switches Error-Down Mechanism.

Why is an interface in Down state on a CE series switch
If the indicator of a connected interface on a CE series switch is off, the interface does not go Up. You can run the display interface command on the switch to view the interface status. Common causes: 1. The interface is manually shut down. 2.The cable or optical fiber is faulty. 3. The optical module is faulty. 4. The local and remote interfaces fail to negotiate parameters including the duplex mode and rate. Recommended solutions: 1. Check whether the interface has been manually shut down using the shutdown command. If so, run the undo shutdown command in the interface view to enable the interface. 2. Connect the interface to another interface, and replace the optical module, optical fiber, or cable on the interface to check whether the interface, optical module, or optical fiber/cable is faulty. If the fault persists, see section "Interface Troubleshooting" in the Troubleshooting - Hardware.

On S series switches, when BFD sessions in Down state receive packets in Up state sent from the peer end, how does the state change
S series switches with their BFD sessions in Down state do not change the status of BFD sessions to Up after receiving BFD Up packets from the remote end. If the link between two communication ends encounters an instant disconnection after the two ends become Up, one end that first detects the link Down event sends a BFD Down packet to the remote end. The remote end sends a BFD Up packet to the local end before receiving the BFD Down packet. The local end does not change its state after receiving the Up packet. This prevents network flapping.

How do I configure M-LAG and obtain information about M-LAG configuration restrictions and examples?
Multichassis Link Aggregation Group (M-LAG) implements link aggregation among multiple devices. One device is connected to two devices through M-LAG to constitute a dual-active system. M-LAG achieves device-level link reliability but not card-level link reliability. For details about how to configure M-LAG, M-LAG configuration restrictions and configuration examples, see M-LAG Best Practices. This document also recommends M-LAG deployment models and configuration suggestions in more multi-feature scenarios.

If you have more questions, you can seek help from following ways:
To iKnow To Live Chat
Scroll to top