Upgrading ePMP from 4.4.3 to 4.5.6 may cause ePMP "halts" a reboot

Products ePMP
1

I had a Force 200 5Ghz in a remote location (as AP and in bridge mode), when I upgraded from 4.4.3 version to 4.5.6 version, the equipment just got lost at the last phase, when rebooting. I have not tried to upgrade from a different version.

I have the idea that it do reboots, but when I try to enter the username and password at the main login page, it seems to halt. But maybe the browser cache showed me the login main page (once).

I replaced the Force 200 with an ePMP 1000 with GPS, and repeated the first upgrade to 4.4.3 version, and configured; two days later I upgraded it to 4.5.6 version, and I lost the radio too. I will have go to the remote location again (at a hill of 4240m high, 2hrs from here, at 40km). Strange thing: the ePMP 1000 with GPS has two firmware banks, the alternate (backup) bank did not worked.

I do remember to be playing with the bandwidth of this new wireless link (from 5Mhz up to 40MHz) before the upgrade, in both “lost” cases.

By the other hand, I had upgraded some Force 300’s, other Force 200’s, other ePMP 1000 connectorized that are in production, with no problem issues, but I did not played with the bandwidth configuration beforre the upgrade.

In my lab, I had to use the “pressing more than 10 seconds reset button switch” method to reset the Force 200, to recover the Force 200 to defaults. Other methods did not worked: First I tried looking for IP’s like 192.168.0.2 or 169.254.1.1 or in the segment it was working or using a DHCP server at the ethernet port and looking for the IP address that could obtain, then I tested disconecting an connecting the POE cable to reset to factory defaults.

Other thing that I have realized is that the LEDs at the RJ45 female port of the Force 200, do not work, but this last thing maybe is a common issue in the Force 200. After the recovery of the Force 200, this two leds did not worked. I will check on a new Force 200 just in case.

I am thinking of any idea to remotely try to recover the ePMP 1000 with GPS (searching 169.254.1.1).

I will inform you.

2

This is the solution I got to recover the remote device.

I could enter to the segment of all the antennas in the remote site:

  1. I issued a ping 169.254.1.1 and I got a response.
  2. I did enter to that device with the browser, using the IP 169.254.1.1, and I did identified that device (IP2, example 192.168.4.10)
  3. I did enter to that device with the browser, using the IP2
  4. I donwloaded the .JSON (text) confg file of IP2
  5. I altered with an editor, the only one ocurrence of 192.168.1.1 address in the .JSON file, with a different one (for example 169.254.1.2). Look for “networkWanIPAddr” tag.
  6. I saved the new altered file: file2.JSON,
  7. I uploaded the altered configuration file, file2.JSON, to that device
  8. I waited about a minute, and repeated this instructions from 1, but, in instruction number 5, I used a different IP for each device.

One of the discovered devices at 169.254.1.1 was the ePMP 1000 with GPS, but it was with factory defaults configuration. So I changed the IP address of that device to the corresponding original IP. I reconfigured using the original IP of that device and bingo !, I could recover the ePMP and the wireless link.

Certainly there should be an alternate way to change the “networkWanIPAddr” IP (using the CLI of SSH with command “config set networkWanIPAddr value”) to each device. But I changed each of the config files of the devices in order to get each IP 169.254.1.1 in the net (segment), identify each device and change that IP to a new one in order to get a new device with the same 169.254.1.1 IP (cambium devices).

Have a look to the next link: