Ethernet Drops Out

spf · Post by **spf** » 01 Sep 2020, 15:15

I have a rather large machine that we were using Raspberry PI 3b+ with Raspbian Stretch running codesys. They performed fine but we wanted a more industrial solution. We purchased 5 of the RevPi core 3+. We have been working with unit serial number 100299 for my testing. When we removed the Raspberry PI and replaced it with the RevPi unit we noticed that the unit kept dropping off the network. We started to ping the unit and sure enough it failed to respond to the ping randomly.

As a troubleshooting effort I downloaded the 2020-06-25-revpi-stretch.img Stretch image from your website and installed it. Next I set the ip address to something common. 192.168.1.131/24. At my desk the ping response was fine. At the machine the ping had drops. Please see image attached.

I then took a wireshark trace of the network. Both the RevPi and my pc were plugged into an un-managed switch. I am not sure what communications are making the RevPi drop packets. I have a wireshark trace but I can't get it to attach to this forum to a provided a image.

I am not sure what the difference is between the RevPi capability and the Standard Raspberry PI 3b+. The Standard Raspberry PI 3B+ has no issues when setup with the exact same IP address and running the same codesys software.

Note that in my testing and the information attached I didn’t load the codesys software or modify the base RevPI image to be sure that I didn’t make a change and cause the issue.

I thank you in advance for your support with this product.

Post by **dirk** » 03 Sep 2020, 15:29

Hi spf, thank you for your good description of the setup. We have analyzed your description and what we noticed is that the Wireshark screenshot seems not to match the ping screenshot.
The wireshark trace shows ARP problems - "Duplicate use of 10.0.0.113!". Could you perform a network scan i.e. with Advanced IP Scanner ?

spf · Post by **spf** » 04 Sep 2020, 19:52

Found the issue. I had purchased 5 units and modified the first one to get it working. Then imaged it to the rest. When I installed them, I installed all of them at the same time. I didn't realize that the image carried the MAC address. So my network had 5 units all with the same MAC address. I fixed that by running "revpi-factory-reset" on each one and the problem is completely resolved. The issue you noticed above with the duplicate 10.0.0.113 address is a Maxcess web guide. There are 10+ on this machine and they have 172.16.x.x address assigned over profinet. The 10.0.0.113 address is some form of PTP verison 1 protocol or discovery protocol. I need to contact them about how to shut that off. Thanks for your help.

Post by **dirk** » 07 Sep 2020, 09:29

Hi spf, never mind, thank you for your positive feedback.