So did you create a LAG with those 3 ports and do you have loop protect and or Spanning-Tree enabled, and do you have a trunk/channel-group or LAG created on the other switch for aggregating these three ports and is loop protect and or Spanning-tree enabled on the other switch? Or do you have separate vlans on each link between the switches?
yes we have separate vlans on each link between the switches, 12 vlans on 3 port. the port whitout LAG config, just trunk. If we enable STP on each vlan on that 3 port, STP should be disable the port. that's mean there are indication loop ?
Why did you put the vlans on separate ports verse creating a LAG? It would simplify any spanning-tree or loop protect issues especially if someone were to make a mistake adding vlan and it a port was to go down it would not affect the link being up but amount of traffic it was capable of would drop. Either way I would check to see if they are supported Ruckus transceivers as mentioned above by Ben and get the info on them.
show media validation
show media eth x/y/z
We have had issues with some Cisco and HP transceivers not working correctly and of course they are not covered by support. What is the other switch you are trying to connect to?