ePMP 2000 kicking us out of web GUI. Tried 2 APs aleady.

ElbowWilham · April 29, 2018, 11:44pm

We are trying to set up a new ePMP2000 with elevated subs. We can not stay logged into the web gui. After we reboot the AP, we can stay logged in for about 10 minutes. After that, we can not log back in. We get the login screen, then after entering the username and password it seems to freeze for a second or 2. Then it gives us a pop-up that says connection temporarily lost.

We also got this error once:

"Device cannot receive the answer, please refresh a page."

We have to reboot the AP in order to get back in.

We have tried 2 brand new ePMP2000s.

When we do get kicked out of the web page, traffic is still passing and there is never any packet loss.

Any suggestions?

ElbowWilham · April 30, 2018, 2:53am

Screenshot

Fedor · April 30, 2018, 8:06am

Hi,

Could you please check continuous ping to ePMP2000 IP address?

What FW version is installed on your ePMP 2000?

Which browser do you use to get access to it?

Thank you.

Cambium_Sri · April 30, 2018, 2:26pm

Hi,

This may also happen when there is an IP address conflict. Please double check that as well.

Sriram

ElbowWilham · April 30, 2018, 3:16pm

I did a continuous Ping and it never drops a packet.

Also, I should note that it never kicks me out of the SSH session either.

Using firmware 3.5.1

I thought about the IP conflict too, so I have tried 5 different IP addresses and it doesn't change anything.

I tried 3 different browsers too. Cleared cache and cookies too.

I seems like some kind of Javascript or cookie bug, but not sure why 2 different APs would do the same thing. We have never seen this on other APs.

ElbowWilham · May 1, 2018, 5:03pm

I found some clients connected to this AP that are throwing off the same error: Connection lost. Please, refresh the browser page.

I switched the IP address with a SM that was not experiencing this error. It didn't make a differenence. The SM is still producing this error with the new IP address. The SM that was not giving me this error is still NOT giving me this error even though the IP address was changed.

So it is not an IP conflict.

Can anyone confirm if this is just a software bug?

Fedor · May 1, 2018, 7:13pm

Hi
I don’t think it is software bug.
We have never seen it in the Lab and other customers have never reported it as well.

Thank you.

ElbowWilham · May 2, 2018, 3:33pm

Thats the only response I get? I can replicate this on 5 different systems now.

You guys obviously programmed that error in right? So what are the circumstances that would cause that error to occur and not let someone back into the Radio?

Fedor · May 2, 2018, 3:52pm

Could you please provide remote access?

You can reach me at fedor.trutsko(at)cambiumnetworks(dot)com

Thank you.

ElbowWilham · May 2, 2018, 5:07pm

I am pretty sure I figured it out and I would classify it as a software bug.

One of our nanobridges had quotes in it's name. When we elevated the bridge, it kept the quotes. Whenever this SM came online, we got the error. I was only able to figure it out because this particular customer is off grid and shuts his SM off overnight.

I guess ePMP does not like quotes in the SM name.

As a software developer, I would consider it my job to ensure my users can not cause these sorts of errors just by injecting quotes. But I also understand it can be an easy thing to over look.

Cheree_King · May 2, 2018, 7:53pm

That must be an Elevated bug because I have a Force 180 with quotes in the name and have had no issues.

Cambium_Sri · May 2, 2018, 11:10pm

@ElbowWilham wrote:

I am pretty sure I figured it out and I would classify it as a software bug.

One of our nanobridges had quotes in it's name. When we elevated the bridge, it kept the quotes. Whenever this SM came online, we got the error. I was only able to figure it out because this particular customer is off grid and shuts his SM off overnight.

I guess ePMP does not like quotes in the SM name.

As a software developer, I would consider it my job to ensure my users can not cause these sorts of errors just by injecting quotes. But I also understand it can be an easy thing to over look.

Hi,

Thanks for narrowing it down. And Cheree King confirming that it works on a Force 180 helps as well. We'll investgate the issue and fix it if indeed turns out to be a software bug.

Sriram

Fedor · May 7, 2018, 8:02am

Hi all,

Issue has been confirmed and internal ticket was raised to fix it.

Quotes in device name cause issue with log in both on SM(Elevate) side and AP side.

Quotes are not allowed for device name on ePMP devices and the only case is UBNT devices elevated to ePMP.

Thank you.

oceron · August 7, 2018, 11:38pm

Hi,

We are seeing this same issue on an AP. The AP only has 1 SM that we elevated recently. The SM is running 3.5.1 and the AP is running 3.5.2. Neither device has quotes in Device Name. Any suggestions would be appreciated.

oceron · August 15, 2018, 3:57pm

I did notice this station has a mac address F09FC252A4B5. Of the 200 + subscribers that we have elevated none of them have that mac address prefix, could this be a factor?

Skullzaflare · August 16, 2018, 2:16pm

We have had a few that do this, normally needing a full power cycle to be able to get into. At which point a reflash usually fixes it.

I will keep an eye out for any "" marks just incase, but we dont use them.