I have running PPPoE servers on few interfaces, so random users cannot connect to some PPPoE servers by the reason of pppoe interface – dynamic, but not running(log says “PPPoE connection from XX_XX_XX_XX_XX_XX was already active” and hungs up, but really users dont connect before, and so the cycling situation for all next connections), same as on topic http://forum.mikrotik.com/t/rb1100ahx2-ros-6-1-pppoe-problem-dynamic-running/67509/1 problem solved by downgrading to ROS 5.25. But another bottleneck – using Intel X520DA2 SFP+ on ROS 5.25 I have freezing whole Ethernet interfaces at random time(( so I cant use that version of ROS. I saw some dependence – freezed pppoe sessions probably hung only at clean Ethernet interface, at same time – if pppoe server running at VLAN on the same physical interface, it’s run ok. Changing MTU of Ethernet interface or PPPoE server does not affect situation. Changing x86 system from Xeon to AMD does not help too.Additional info – I use radius auth for ppp.
I’m not only one!
Any advance, MT team?
I think this is not a x86 problem but 6.x version at all.. I have same problem at x86 machine with Core i7 CPU and CCR1036-12G-4S.
On 5.25 all works fine…
Have the same Problem on rb1100 AHx2
Have the same problem with 6.0 6.1 and 6.2 on 1100AHx2 and 433AH and 439G . Mikrotik reckons there is no issue
I have same issue on my several CCR1036. It`s known bug on V6.2. They said bug fixed on 6.3. Unfortunately 6.3 has own brand new bugs. 