We have installed about 120 ePMP 3000L APs and lately in completely random times they just stop responding and deregister all SMs connected. The ethernet cable remains UP (it never goes down/up) but the AP is no longer accessible (dead), not even responding to 169.254.1.1. The only way to solve is to power off and power on again. This is happening with any of the latest firmwares up to 4.6 and to multiple 3000L APs in completely different places of our network. Until now I had to power off and on a random 3000L AP at least 30 times.
Does anyone else experience this problem with the 3000L?
This never happened with an 1000 or 2000 AP ever before. What is happening with the 3000L APs?
1 Like
Hello,
Have you opened the support ticket?
Did you observe if it is related to low temperature?
How often the issue occurs on a single ePMP3000L AP?
Thank you.
Hello Fedor,
I opened a support ticket this morning. I attached tech support file but I am not sure what you can find there since I have to power off and on to make it work again.
It is not related to low temperature. Last days temperature doesn’t drop under +5 Celsius at night here.
It is completely random, to some it can happen after some hours while others had uptime even for a month. Example, yesterday afternoon a had to power off/on one and just about 6 hours later blocked once again. Two days ago another one that had uptime of 30 days. One week ago another with uptime of 90 days.
I would be grateful if you could send me the support ticket number.
We would propose performing RMA for 1-2 most affected APs to grab them back for the investigation.
Since as you mentioned in that case we don’t have many options for the debug when the device is not reachable nor from Ethernet either Wireless.
Thank you.
I sent you a PM. Thanks for the support
@someone Have you noticed if this blocking is happening more on higher loaded sectors or one with high subscriber counts?
Thanks!
It is happening even with 0 subscribers connected
Reading about the problem there was a click as if it had already been seen a few years ago.
Proxim Tsunami MP11 had integrated peltier modules with which they maintained the internal temperature.
Great idea but someone in writing the firmware was careless and the device did not distinguish between faranheit and celsius degrees.
There were similar problems with that, especially in a temperate climate.
Not at all just at some
greeting
I’ve noticed this as well. Let me know what happens on this case.
Not good. I have been debating using 3000L or 2000 for some AP’s on a new tower. Getting close to deciding. I wanted to use the 3000L for additional capacity but keep seeing more issues. About the time it looks fixed, a new one popps up.
To be fair – the people not having issues, aren’t posting for support in the forums. 
I mean – we have a bunch of 3000L’s in the air (and I have my gripes with Cambium business decisions) but I don’t have any performance issues or GPS issues, or lockup issues at all with 4.6.0 - Zero.
There are WISP’s with literally hundreds of them in the air, and if we’re not having issues, we don’t post that we’re not having issues. Except this post I guess… LOL.
2 Likes
Yo tengo un 3000L y tambien le ocurre lo mismo
@ninedd Thanks for the positive post. I think I will go ahead and try the 3000L.
This tower won’t have a lot of SM’s but several will be business so I want to be able to offer higher bandwidth. The 3000L specs look perfect.
1 Like
Someone,
have you turned smart speed off? I have found this feature to cause too many problems as not every switch uses this feature properly.
Following – we have the same thing happening on one of our units.
I also tried disabling smart speed but nothing changed in this case.
I sent two of our 3000L with the most frequent blocks to cambium for further inspection.
Hello, we are facing the same issue on several sites.
We are doing testing in the lab on the provided devices.
We will update you once we have more information.
Thank you.
What switches take issue with the smart feed?, curious…
This is just our observations and I do not expect yours to be the same nor even agree with ours.
Smart speed is an intresting creature that when it works properly its great, but this is rarely the scene. So far I havent seen a single switch from any brand actually work 100% across differing switch brands and devices.
We have cisco, hp, palalto, dlink and netgear switches and thes all play well with their own brand, but god forbid you want an hp switch to use smart speed to a cisco.
This is exacerbated with bridges, media converters and even dumb l2 (though smart speed capable) switches.
The problem isnt the technology but the independent implementations. Each manufacturer does things their own way and its not always understood on the other end of the cable. Eg cisco switch decides to drop from 1Gbps to 100Mbps but doesnt drop the carrier link. Theoretically the whole link should just drop speeds and carry on like nothing happened, but in reality the link halts due to speed mismatch until you drop the link and bring it back up. Now add a wireless bridge which provides no concept of link state monitoring from switch to switch, you have to rely on the local radio connection to renegotiate and this doesnt happen properly.
In short, it would be a great tech if everything actually worked properly but since it is literally only a few years old the bugs and compatibility havent been worked out.
If we leave it on we have a lot of calls for internet outages, simply disabling it has solved so many issues that point to ports and cables.