The S105 switch at a certain office is repeatedly up/down when docking with third-party device aggregation member ports

NULL

On site aggregation group BAGG9 has two member ports (1/0/0/9 and 2/0/0/9) for link aggregation with third-party devices on the opposite end. This end is a dynamic aggregation, and the selected states are all normal. However, it was found in the logbuffer that there is a regular up and down of ports 1/0/0/9.

Observing the updown records, it was found that the updowns were switching in seconds and very frequently:

%Mar 16 07:05:48:598 2019 HEXIN LAGG/6/LAGG_INACTIVE_PARTNER: Member port GE1/0/0/9 of aggregation group BAGG9 changed to the inactive state, because the aggregation configuration of its peer port is incorrect.
%Mar 16 07:05:48:620 2019 HEXIN IFNET/5/LINK_UPDOWN: Line protocol state on the interface GigabitEthernet1/0/0/9 changed to down.
%Mar 16 07:05:55:595 2019 HEXIN LAGG/6/LAGG_ACTIVE: Member port GE1/0/0/9 of aggregation group BAGG9 changed to the active state.
%Mar 16 07:05:55:632 2019 HEXIN IFNET/5/LINK_UPDOWN: Line protocol state on the interface GigabitEthernet1/0/0/9 changed to up.

Upon checking the configuration of the peer device, it was found that the peer device was configured with a short timeout, as shown below:

Our device defaults to a long timeout. Inconsistent timeout times on both ends result in frequent UP/DOWN on the local port. 

The LACP timeout refers to the timeout period for a member port to wait to receive LACPDU. After the LACP timeout period, if the local member port has not yet received LACPDU from the opposite end, it is considered that the opposite member port has expired. The LACP timeout also determines the rate at which the peer sends LACPDU. LACP timeout has two types: short timeout (3 seconds) and long timeout (90 seconds). If the LACP timeout is short, the other end will quickly send LACPDU (1 LACPDU sent every 1 second); If the LACP timeout is a long timeout, the other end will slowly send LACPDU (sending 1 LACPDU every 30 seconds).

Adjust the LACP timeout time of our company's equipment to a short timeout, or adjust the peer equipment to a long timeout.