Ethernet dropped out on Connect +

[dirk]

Hello Chris, I have received your device. I could successfully login. This are the tests I have performed
I have connected my laptop via

• USB-Ethernet to the RevPi
• the built in Ethernet interface to the RevPi
• the built in Ethernet interface and a switch to the RevPi

Each time

• I started a SSH connection.
• Then i have switched off the Ethernet A interface from the RevP
• I have waited a few seconds
• re-plugged the Ethernet cable
• Restarted the SSH connection

Unfortunately I was not able to reproduce the erroneous behaviour.
I have had to re-start the SSH connection but it was always successful.
At the end I have also used HDMI and an USB keybard. Logged in and pinged web.de, disconnected the cable and connected it again and the pings did work again.

Could the error result in something that is involved in the network communicaton on your side?
What is the topology and the devices that you use?

[ChrisDeveloper]

Hi Dirk,

Did you manage to recreate the ethernet drop out issue?

Kind regards,
Chris

[dirk]

Hi Chris, no unfortunately I was not able to reproduce the error.

[ChrisDeveloper]

Hi Dirk,

Sorry I didn't see your previous post! Here is the network topology.


Main Controller PC (Ubuntu 20.04)
2 ethernet interfaces:

-> 192.168.1.x
- Office Network

Seperate network with only modbus TCP/IP devices.
-> 192.168.2.x
-- Kunbus Connect+ : 192.168.2.3
-- Kunbus Core3 : 192.168.2.4
-- 2 * Modbus devices TCP/IP

The error happens after a few days, could I suggest using the suspect Kunbus Connect+ to talk to another Modbus device TCP/IP and check back every couple of days to see if the ethernet connect is still working? This test scenario would be very similar to our workshop.

Kind regards,
Chris

[dirk]

Hi Chris, I have some updates on this case.
I have made a test setup with a RevPi Core 3+ device and your RevPi Connect+ device.
Core3+ = Modbus TCP Slave
Connect+ = Modbus TCP Master

Both devices with kernel version:

Code: Select all

4.19.95-rt38-v7+ #1 SMP PREEMPT RT Tue, 16 Jun 2020 13:25:43 +0200 armv7l GNU/Linux

The Modbus TCP Master is configured as follows:

Modbus TCP Master Extended Data Configuration

2020-12-10 15_39_34-Window.jpg (36.98 KiB) Viewed 2304 times

On the slave I wrote some data

Code: Select all

piTest -w Output_1,65

On the Master I have tried to read the value ...

Code: Select all

piTest - r Input_Word_1

... but I received "0".

The output of

Code: Select all

piTest -r Modbus_Master_Status

reported "255". This was an undocumented error.

So I have checked everything twice, pinged the devices, checked the configued Slave-IP Address on the PiCtory Modbus TCP Master, also

Code: Select all

tail -f /var/log/daemon.log

At the end I have updated the devices with

Code: Select all

sudo apt update
sudo apt upgrade
sudo reboot

and the communication worked out of the box (!).

Just to mention it - I have had to comment out an apt proxy that was configured on the Connect+ device, here:

Code: Select all

/etc/apt/apt.conf

So at the end I would say we have a situation that was not clear before we decided to do a RMA case here.
If you don't mind I'll ship the device back to your hands and you can verify if the erroneous behavior is fixed with the latest updates.