Other potential reasons are listed below:
- If there is a Service Provider link between two related nodes this should be checked as well.
- Due to ESXi server that was sending pause frames to the edge device, slpp kicked in and brought down the links ERS edge device was configured in lossless mode.
"Lossless mode is mostly a mechanism for implementing symmetric flow control in a modern switched Ethernet environment. Most vendors support asymmetric flow control by default. With asymmetric flow control, the switch will honor the pause frame in that it will pause traffic to the device that sent the pause frame but it will not forward that pause frame. The result is that the server is protected from being overloaded but the senders are not notified to scale back its traffic and upper layer protocols need to handle the back-off and re-transmission."
- Port flaps might be observed due to VLACP PDU missing from the remote end, hence it will automatically shut down its own port assuming that the other port was down. Default timer is set to 500 which is good time for exchange, however we can increase the fast periodic timer to 1000 to avoid such occurrences.