v7.1.1 RB5009 Link SFP+ <---> any device with SFP (SFP+) packet loss 35-50%, regardless of the load. Tested on two RB5009.
v6.49.2 CRS328 Link SFP+ <---> any device with SFP (SFP+) packet loss 0%, regardless of the load.

Everyone has this problem with packet loss on the port SFP+?

v7.1.1 RB5009 Link SFP+ <---> any device with SFP (SFP+) packet loss 35-50%, regardless of the load. Tested on two RB5009.
v6.49.2 CRS328 Link SFP+ <---> any device with SFP (SFP+) packet loss 0%, regardless of the load.

Everyone has this problem with packet loss on the port SFP+?

Are you able to let us know a little be more about how you are testing the devices so others might be able to replicate? More information about how the router is configured, what SFP module you are using, how and where you are generating the traffic, where the traffic is flowing to etc would be greatly appreciated

i have some issues too with this port. i connected this port with an unraid server (Mellanox ConnectX-3) and try different 10G modules with maximum 2 meters cable. The connection between this adapters are succesfully every time i start the computer or the RB5009.
But after a time, and this time is different (2h, 4h, 30minutes, 22h etc..etc..),it will be lost the connection. after a reboot it works again. i have tried so much but i cant solve this problem. if this problem are a MT RB5009 problem or a problem with the Mellanox Card i dont know. But before i use the mellanox cardwith other hardware it works every time well.
this is not a problem with the throughput rate because sometimes there are no data goes through the adapter but the connection connection will be lost.
maybe this is the same problem like you have.
now i connected the unraid server with an 2.5G connection (Ethernet RJ45) to the 2.5G MT RB5009 port and it runs and runs and runs....

can anyone help ???

I have the same, trying with different SFP 1G modules (Mikrotik, Tplink, Ubiquiti) and all of them when the fiber is longer than 5meters fails the negotiation and many link downs at RB5009 side.

With a fiber of 50Meters, one side there is hEX POE RB7.1.5 with S-31DLC20D (auto-negotiation=no speed=1Gbps full-duplex=yes) other side RB5009 RB7.1.5 with same S-31DLC20D (auto-negotiation=no speed=1Gbps full-duplex=yes).

hEX POE shows link up, but the RB5009 log:

sfp1_SW link up (speed 1G, full duplex)
sfp1_SW link down
sfp1_SW link up (speed 1G, full duplex)
sfp1_SW link down
sfp1_SW link up (speed 1G, full duplex)
sfp1_SW link down
sfp1_SW link up (speed 1G, full duplex)
sfp1_SW link down
....
every second the same.


This same setup worked for years hEX POE with RB760, but we decided to upgrade to RB5009... :-/

ROS v7 is being actively worked on, so make sure the problematic device is running the latest from the series ... in case of 7.1 that's currently 7.1.5. Quite a few users have good experience also running 7.2rc5 ... which is still development version (release candidate), but some things are done much better than in 7.1. And regardless the marketing name (stable, development), any ROS v7 still has many teething problems, so if one has problems when running 7.1 there's no reason not to try 7.2rc. Just make sure backup and text export of configurations are created before upgrading (and downgrading) ... just in case if roll-back is needed.