ePMP 1000 now starting to cause a LOOP

Hello everyone,

I don't know if anyone has experinced this problem as of lately, because now we are starting to see that our ePMP 1000 SM's, cause some loops on our network. So far what fixes the issue for us is, we have to shut off the Ethernet on the SM, reboot the SM and that kind of fixes the problem (in most cases). But, this doesn't always fix them and turns out the radio is bad, after we replace the radio, no more loop errors. 

Has anyone else experinced this problem? and might know any other solution to fix these loops with out having to replace the whole radio unit? Seems like this happens when the radio went through a bad storm but still works you know and it looses its mind. 

Also we have the SM's on Bridge mode! for a heads up, if you are wondering what configuration mode we are using. 

If the radios are in bridge mode how could they cause a loop ? Wouldn't it be the device behind the radios ?

Also, what do you mean when you say they are causing a loop ? That doesn't seem possible for an end device to make up very specific data that just happens to correspond to actual data on your network and then do it in such a way as to cause a loop. 

We see the same thing when a radio is damaged, HOWEVER since moving away from a bridged network, it only affects everyone on that panel. Generally we notice via packet loss to all the SM's on that panel.

Same resolution however. Disabling eth port fixes it until the radio can be swapped.

Happens with both NAT and Bridge SM configs. 

1 Like

@brubble1 wrote:

If the radios are in bridge mode how could they cause a loop ? Wouldn't it be the device behind the radios ?

Also, what do you mean when you say they are causing a loop ? That doesn't seem possible for an end device to make up very specific data that just happens to correspond to actual data on your network and then do it in such a way as to cause a loop. 


When the SM eth port/chip is damaged, sometimes it "mirrors" back to the network and makes that range think there is another path to go. Atleast thats how it appears in tower router logs

1 Like

Yes we have seen this, and yes replacing the radio is the only solution we have found.

How can we detect via logs…any specific error message?

considering this is a necro-post of three years, you may want to start a new one and detail your issue as clearly as you can.

loops are normally found by either broadcast storms about 2 mins after you plug something in or if in the event a radio is reflecting packets back into the network, with your switch logs which will show the behavior and packet returns on same port. Spanning tree reports are also a good way to find un-intended loops.