I have an issue with several RB3011 boards.
Some ethernet ports are often flapping. For example, they go from 1 Gbps to 100 Mbps negotiation, or they disconnect unexpectedly for short times (1 -3 seconds).
Of course this behaviour is causing lot of problems (I see frequent OSPF flapping between routers).
On the eth stats I often see some FCS errors.
I already checked my cat5e cables and plugs with Fluke.
The interesting thing I noticed is that the disconnections seem to be present on all ports of the same switch chip at the same time.
Please look at the following example, where you can see that eth6-eth10 flapped at the same time (they are connected to different devices but share the same switch chip internally)::
I tried with different RouterOS, now I use 6.42.10 (longterm).
I’m on mobile but search this forum for the term “port flopping”. There is a large thread about it, why it is happening and how the problem hasn’t yet been fully solved.
I had to add a gb switch and move all my connections to it, which solved the flap. I did notice that the flapping disappeared when I used ONLY gb connections on ports 2-5 ONLY. Using a switch improved my network 10x and I am now reconsidering Mikrotik routers as the best since this problem is known and has been known for months and Mikrotik even released the RB4011’s knowing they has the same issue and even worse the RB4011 SFP+ port is a red headed step child and only works if you can figure out the proper combination of distant end device and sfp type installed.
But it’s really sad you have to use additional switch to get features that should work on RB3011 itself, only because someone decided to save like $1 BOM cost on switching chip in RB… and this mistake was never fixed and RB3011 is still being manufactured with this flaw today.
Guys need help Setup:
3 ISP + Router + MIkrotik(Clients: PPOE Radius) + OLT to clients Optical Network Unit (ONU) the problem is that most of our clients keeps on disconnecting.
