I'm experiencing an issue since upgrading our devices to 7.20. It seems to be happening on the lhg60 devices, but may have occurred on other microtik devices as well, we only just started monitoring as we were seeing the same issue pop up.
Basically a bridge (only bridge) on the device stops passing traffic. The fix is easy as we just create another bridge and move the ports over to the new interface. The new interface works so we delete the old interface. Although an easy fix it does cause service interruption to our clients. It also happens every 2-3 days.
Again, just wondering if anyone has seen this or if mikrotik is aware of it. Also wondering if there is a permanent fix if it has been noticed
Jonathan
If you can, please export support.rif file while your devices are working and export one after they failed and sent those files to mikrotik support so they can check if there is a bug.
Good Idea… I’ll send that off now.
Sent off the Supout.rif files to support and got a response 4 days ago that they are looking into it. Anyone have this happening and anyone think of any solutions. Our Customers are getting a little annoyed.
Why stay on 7.20 if it is such a problem ?
EDIT: fixed autocorrect typo
Assuming it’s why stay on 7.20. If downgrade goes wrong I don’t have access to these units. Brief interruptions (however annoying) are better then a couple of days outage
If you’re able to create the new bridge and move over the ports, the device is accessible ?
I'm curious, how can that be ? How does the setup look like ?
Well, you have to prepare yourself that you will probably wait long for Mikrotik to get back to you… And then probably you will wait more until problem is resolved...
So as @holvoetn said, best option is to roll back the ROS version to one where everything worked.
The bridge only bridges a vlan between the wireless interface and the ethernet interface. The ethernet and wireless interfaces themselves have IPs and route between them.
We found if we try to route traffic between wireless and wired we can’t reach gigabit speeds as there is not enough cpu power, so we just bridge the vlan that need 1gbps speed. Management is outside of that.
So when the bridge stops working it stops passing traffic down that vlan but other traffic keeps flowing
One other note is it seems to be only on lhg60 units. We aren't seeing this on any nrays
Whoops…. In my last post I said it isn’t affecting any nrays. I’m wrong, it is. Just went back and looked.
And I should also be specific on setup. The bridge in question consists of 2 ports that are vlan interfaces bound to the ethernet and wireless interface respectively.
The below is cleaned up config from a unit
/interface bridge
add name=bridge1658
/interface bridge port
add bridge=bridge1658 interface=vlan1658w1
add bridge=bridge1658 interface=vlan1658e1
/interface vlan
add interface=ether1 name=vlan1658e1 vlan-id=1658
add interface=wlan60-1 name=vlan1658w1 vlan-id=1658
You need to set protocol-mode=none on the bridge in this kind of setup. See Layer2 misconfiguration - RouterOS - MikroTik Documentation.
Hmmm…interesting. Ok, changed the protocol from rstp to none and will see if that works
OK…. Changing the protocol to none seems to be working (at least the issue hasn’t happened since the change was done. Don’t necessarily agree with it as I’ve seen people create loops with vlans a lot more often then with physical interfaces, but if it keeps traffic flowing, then what can I say.
Also recieved an email from support that the problem is fixed in V7.20.2. Hopefully it’s released soon
FYI for anyone experiancing this. V7.20.2 is out and the stp fix is at the top of the changelog. Thanks for the help while I was waiting for this.
Jonathan