Hello,
I am fairly new to working with Artiq but recently we encountered a problem where we could not connect to the Kasli core device over ethernet, did not receive any response when pinging it. We also noticed the red error led was on. Connecting to the core device via USB gives us following log:

I simultaneously tried to reconfigure the ip address with artiq_mkfs which got rid of the red error led but the problem with the lost SDA arbitration still persists.

    NLack I simultaneously tried to reconfigure the ip address with artiq_mkfs which got rid of the red error led

    What does that mean?

    The ERROR LED is supposed to light up every time there is a "panic" (in the Rust sense of the term), such as this I2C issue.

    What do you have connected on the I2C bus?

      sb10q

      Hey,
      at first I thought that there was a network problem as I did reconfigure out local network and thought that I might have messed something up while doing so. So I wanted to make sure that the IP address was not changed or similar, but the error LED came back anyways.

      In the meantime, I disconnected the individual modules by pulling the ribbon cables of the Kasli to see if any module causes the error but nothing so far. We have following modules connected:

      1x Clocker
      3x SMA TTL
      1x Grabber
      3x Urukul 9910
      1x Sampler
      1x Zotino

      Take a look at the I2C signals with an oscilloscope. It's also possible that the I2C switch chip has been damaged e.g. by ESD.

      Kasli hardware version mismatch is also a possible hypothesis for that symptom.

      I picked up the I2C signal from the patches on the board, and they had a constant 1V on them (both SDA and SCL). I am not sure if the I2C has to be triggered, regardless I tried running the test_i2c.py but got no results.

      Not sure if I understand correctly, but the box worked for months before, so there are no hardware or software changes.