ePMP Software Release 3.0.1 is now available

ePMP Software Release 3.0.1 is available at: https://support.cambiumnetworks.com/files/epmp/

Does this address the DFS issue that was introduced in 3.0? I don't see any mention of that in the reelase notes.  

http://community.cambiumnetworks.com/t5/ePMP-2000-and-1000/DFS-broken-in-3-X-beta-and-production/m-p/58600#U58600


@Au Wireless wrote:

Does this address the DFS issue that was introduced in 3.0? I don't see any mention of that in the reelase notes.  

http://community.cambiumnetworks.com/t5/ePMP-2000-and-1000/DFS-broken-in-3-X-beta-and-production/m-p/58600#U58600


Unfortunately no. We're still working on that. Current target is the next release which is 3.1. 

Next release?

I'm quite irritated over this. First, you knew about the issue and released a new version, and didnt even put this known defect in the "known issues" section, its been more than a month. This is a major bug, not a minor incovenence with a workaround.

We have had to start transitioning our customers over to Mimosa gear because we have some *&^** head running Ubiquity gear at full bore on the non DFS 80 mhz channels less thn a mile away from up with his gear pointed in our direction at the same elevtion as our tower. Now wen hes in his system randomly flipping channels he knocks us offline.

After our transition we will be having a fire sale to get rid of our Cambium gear.

While I so feel your frustration, I’m in the market for some ePMP 1000 GPS APs at fire sale prices!

We’ve left our DFS APs in a 2.6 release and they have been fine. We did have some angry customers though before we figured out why nothing would stick on DFS in 3.0. Not something you expect to happen. It certainly should be listed as a known bug.

1 Like

Hi MarioL

I apologize for the trouble caused due to this defect. False DFS detections are very challenging to resolve as we have to make sure we are not missing real radars while trying to ignore false pulses. This particular issue involves pulse widths of type 1 FCC radars which is very close to widths of frame based WISP radios. So it makes it that much harder to filter out the non-radar pulses while staying legal. We have been working on a solution and hope to have something in the next couple of weeks as part of the 3.1 beta release. As AU Wireless said, you can always fall back to Release 2.6.2.1 to avoid this. The DFS detection mechanism in 3.0 had to be changed to accomodate the ePMP 2000 platform and it resulted in the radar detection mechanism being sensitivie to non radar pulses. That is what we are trying to solve as a high priority. 

We released 3.0.1 as it has some much needed enhancements and bug fixes for ePMP 2000 customers and we didnt want to delay it any further. That said, our goal is always to have a release that is of high quality for all platforms and all customers. 3.0.1 is an exception. 

I appreciate that running your network with the best gear available is a priority for you and I understand you wanting to transition over to another vendor. But once we resolve the DFS issue, I hope you will give us another chance. 

AU Wireless, Yes, the DFS issue must absolutely be listed in the release notes. It was a miss on our part. We will have an updated version of Release Notes on the support web-site this afternoon. Just FYI, the DFS bug tracking number is 12329. 

Thanks,

Sriram

1 Like

We did try to revert back to the previous version, but bugs in inte CN system led us to abandon it, also the 2.6.2.1 had bugs in it and cause many of our SM's to flap that were fixed in the 3.0 version. Because of the "stuck"  firmware processes that were never removed by support in CN and false outages we just moved on to a system that has far less issues. We grew tired of whack a mole, and constantly having to flash sevices to figure out a balance to keep things up.It felt as though we were running a system on the edge of disaster because of the unstable firmware. We are trying to run a business and feel that Cambium is to flakey for us to use. We spent a lot of money on a system that looked good on paper, but in our enviroment turned out not to be the case. Ive had people suggest moving to the 200 series,, but Im not willing to throw money at a dead horse, We bought this system apparently just before the 2000 series came out.

Mario, why don't you run 2.6.2.1 on JUST the AP, and keep all the clients on 3.0.1 stable?

1 Like

Tried updateing a cuple SM's to 3.0.1. Found another new bug. In wifi mode, setting a prefered AP, the password box is converting non text number caracters to ASCII codes on save, and reboot. causing a truck roll to log into the force 180 so you and re-save the password.

Hi,

We cannnot reproduce described issue.

I would be really appreciate it if you could provide detailed steps to reproduce.

Thank you.

The new version of the  release notes has not made it to https://support.cambiumnetworks.com/files/epmp/.  

Issue 12329 IS NOT listed in "Known problems or limitations (System Release 3.0.1)"


@MarioL wrote:

Tried updateing a cuple SM's to 3.0.1. Found another new bug. In wifi mode, setting a prefered AP, the password box is converting non text number caracters to ASCII codes on save, and reboot. causing a truck roll to log into the force 180 so you and re-save the password.


Are you adding the preferred AP under Configuration->Radio or under Monitor->Wireless?

j


@system Administrator wrote:

The new version of the  release notes has not made it to https://support.cambiumnetworks.com/files/epmp/.  

Issue 12329 IS NOT listed in "Known problems or limitations (System Release 3.0.1)"


Hi, 

Its listed under "Known problems or limitations (System Release 3.0)" because customers have reported it in Release 3.0. 

Thanks,

Sriram

Yes it does not like the < caracter

Any ideas why the active antenna is constantly changing?  Prior to this release it was much more static and not dynamically changing all the time.  It could be that its performance is better now but it just seems for fixed wireless connectivity that is constant distance, angle, etc that it would not select different antennas quite so often.  

your thoughts?

FYI - those screen shots were about an hour apart....

Hi,

In 3.0.1 several antenna selection algorithms were added.

AP every 10 minutes perform verification whether the best angle and antenna was selected before.

And you see different angles values and antennas selected because of environment is not stable and for different moments of time interference has different direction and power level.

With 3.0.1 ePMP2000 with Smart Antenna provides better throughput in noisy environment then in 3.0.

Thank you.