MikroTik

Recently upgraded six RB1200 routers to v6.11. Three went from v5.26 and three went from v6.7.

Now, we experience unstable ethernet connectivity to Ubiquiti radios (NanoBridge and RocketM5). I am seeing negotiated link speed fluctuating from 100/FULL to 10/HALF to 10/FULL. Hard setting the radio and router interface to 100/FULL does not help. Radios on v5.5.6 or v.5.5.8 - both experience the problem.

When the issue occurs the port continues to show "R"unning state but any services running on the interface stop - EoIP/OSPF/RIP etc...

To fix, you need to power cycle the radio and then disable and enable the router interface. It fixes it for a random amount of time. Sometimes minutes and sometimes hours...

Issue confirmed here: http://community.ubnt.com/t5/Installati ... d-p/793182

This problem is not occurring where we run v6.10 of RouterOS.

Is this fixed in 6.12?

Thanks,
Scott

Have you upgrade the firmware?

Look to "system" "routerboard" click on "upgrade" if not upgraded

Yes. RouterOS firmware updated.

Try with routeros 6.12 ore ask to Mikrotik 6.13 rc

Thank you for your replies, but I'm not jumping on the "upgrade firmware" bandwagon anymore.

Per the release notes for v6.11 it was indicated that TX/RX Flow Control was added to ETHERNET. Obviously there is something more happening here which is affecting link state with the radios.

Thank you for your replies, but I'm not jumping on the "upgrade firmware" bandwagon anymore.

Per the release notes for v6.11 it was indicated that TX/RX Flow Control was added to ETHERNET. Obviously there is something more happening here which is affecting link state with the radios.

More and more it's essential to have a test section on a network (regardless of which manufacturer product is being used?) before upgrading live sections of a network.

@n21roadie - Isn't that the truth now, isn't it!

I'm hoping some other people chime in here about this.

@n21roadie - Isn't that the truth now, isn't it!

I'm hoping some other people chime in here about this.

As I said before on many posts - I don't subscribe to "Newer is better" until the version is proven to be better?

I upgraded all my RB450G from 5.26 to 6.12 and I don't see any ethernet issues.
They're connected to NanoBridges and Rockets.

Inspect your network cables if the are good.
Ping from the routerboard to the ubiquiti device en look to the ping time ore the are stable.
If the are not stable replace the network cable

Recently upgraded six RB1200 routers to v6.11. Three went from v5.26 and three went from v6.7.

Now, we experience unstable ethernet connectivity to Ubiquiti radios (NanoBridge and RocketM5). I am seeing negotiated link speed fluctuating from 100/FULL to 10/HALF to 10/FULL. Hard setting the radio and router interface to 100/FULL does not help. Radios on v5.5.6 or v.5.5.8 - both experience the problem.

When the issue occurs the port continues to show "R"unning state but any services running on the interface stop - EoIP/OSPF/RIP etc...

To fix, you need to power cycle the radio and then disable and enable the router interface. It fixes it for a random amount of time. Sometimes minutes and sometimes hours...

Issue confirmed here: http://community.ubnt.com/t5/Installati ... d-p/793182

This problem is not occurring where we run v6.10 of RouterOS.

Is this fixed in 6.12?

Thanks,
Scott

Issue is still prevalent.

I downgraded from 6.11 to 6.10. Replaced the ethernet cable. Changed ports on the router. Changed ports on the PoE switch. Changed the patch cable. Changed the radio.

Even went as far as disabling the EoIP tunnel on the interface.

Ran great for seven hours and went down again.

When it goes down you can see the radio from the remote end but it loses connectivity with the router at the root end. Interface shows running state. You have to power cycle the radio and then cycle the interface. Comes back up instantly.

This router has two UBNT radios attached to it. Only have issue with one. Like I said the radio has been replaced.

Next week im dropping a different router in.

Scott

ISSUE RESOLVED

I pulled the RB1200 and installed an RB1100AHx2 running RouterOS 6.10 and firmware 3.10 <--- SOLUTION

No issues for over 48 hours.

I will edit and mark solved once I feel the issue is gone.

Has anyone found the fix for this, the firmware is now at 6.24 and we are still unable to move past 6.10 because of this bug with Ubiquiti gear connected to. We have tried Ubiquiti firmware 5.6beta6 as well and still the same results. We have only seen this issue on the RB-1200 and the quickest way to get it back online is reboot the whole switch.

@wispvt

As stated in my post, I resolved this by replacing the 1200 with an 1100.

I will also point out that this issue started to occur on more of my 1200's over the last six months. In every instance we just swapped with 1100's. Furthermore, the majority of the routers that did have the issue were located in colocations with other providers gear mounted close by and other providers wiring running with ours. I did read at one time that the 1200's are prone to some sort of cable interference.

Scott

There has to be something with the 1200's and the firmware past 6.10. I hate to throw out the 1/2 dozen 1200's we have due to obviously a firmware issue. All our other switches are fine. I have kept hoping it would get fixed in each version but to no avail. It would be nice if someone from Mikrotik would take a look at the issue as it is easy to replicate and a lot of Ubiquiti users use Mikrotik for their switches.

Currently facing the same issue ethernet link from Radio locks up showing zero rx in the Mikrotik Router...

my experience...
-new cat6 cables
-new cat6 connectors
-new poe switch (previously PowerBoxPRO, now netPower 16P with dual power)
-Different ports
-Different Rocket M2 (new)
-Autonegotiation disabled / enabled
-Fixed to 100HalfDuplex
-Now in automatic

-Same problem...
Aug / 22/2020 19:50:05 interface, info ether7 link down
Aug / 22/2020 19:50:08 interface, info ether7 link up (speed 100M, full duplex)
Aug / 22/2020 22:55:47 interface, info ether7 link down
Aug / 22/2020 22:55:49 interface, info ether7 link up (speed 100M, full duplex)
Aug / 23/2020 01:19:48 interface, info ether7 link down
Aug / 23/2020 01:19:51 interface, info ether7 link up (speed 100M, full duplex)
Aug / 23/2020 18:56:38 interface, info ether1 link down
Aug / 23/2020 18:57:54 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 18:58:30 interface, info ether1 link down
Aug / 23/2020 18:58:33 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:00:29 interface, info ether1 link down
Aug / 23/2020 19:00:34 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:01:03 interface, info ether1 link down
Aug / 23/2020 19:01:07 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:01:33 interface, info ether1 link down
Aug / 23/2020 19:01:38 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:02:07 interface, info ether1 link down
Aug / 23/2020 19:02:11 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:15:30 interface, info ether1 link down
Aug / 23/2020 19:15:35 interface, info ether1 link up (speed 1G, full duplex)
Aug / 23/2020 19:16:04 interface, info ether1 link down
Aug / 23/2020 19:16:08 interface, info ether1 link up (speed 1G, full duplex)
Aug / 24/2020 06:18:20 interface, info ether7 link down
Aug / 24/2020 06:18:22 interface, info ether7 link up (speed 100M, full duplex)
Aug / 24/2020 15:18:13 interface, info ether7 link down
Aug / 24/2020 15:18:15 interface, info ether7 link up (speed 100M, full duplex)

we have random flapping or bad negociation (100m instead of 1g) on our 4 new netpower. running from 6.46 to 6.48 fw version. Is it a known issue?