Reachy2 SDK does not connect

Users Support
1

Hello,
I am having a problem with Reachy2 where we cannot seem to establish a connection to the SDK. In python, we receive the error message “Could not connect to Reachy with on OP address , check that the sdk server is running and that the IP is correct.”. However, the dashboard suggests the servers are all running successfully.
Similarly, if we access the “Robot” tab in the dashboard, we get a warning that says “Cannot connect to the robot with the Python SDK. Make sure that the core service is running.”
If we attempt to connect via the teleoperation pathway, we can only get as far as the transition room. The status information says “Connection to Reachy has failed” and warns that the network connection is slow.
We have tried many things, including rebooting, reinstalling software, updating and downgrading Reachy’s software via PLUM, connections with various different devices and networks. I have previously (circa May 2025) succeeded at establishing connection with the SDK, but not since. There was also one instance, similarly early on, where I was able to see an image from Reachy’s camera frozen in the transition room view before the connection failed.
We would greatly appreciate some assistance. Is this a familiar issue? Is there a way to “factory reset” Reachy2 and please reattempt to use/connect from a clean slate? Or a known set of versions for the server software that we could revert to that should work? Any information would be greatly appreciated.
Thanks.

2

Hi @hharland ,

Could you restart fully the robot i.e.:

  • reboot the bedrock (Reachy’s computer)
  • press the red emergency button
  • press the blue lighted round button on the mobile base
  • unlock the emergency button
  • press again the blue light round button
  • turn on Reachy’s computer

and send the logs of Reachy’s main service?

You can get them by doing

journalctl -b -u reachy2-core > log.txt

in a ssh session on the robot and then get the logs in your local machine using wormhole or the scp command for example.

Also could you check on PLUM for each service the version of the service installed on your Reachy and send them?

Thanks

3

Hi Simon,

Thanks for the reply. I have uploaded the logs here (from an initial complete shutdown). Please let me know if there is a better way to send them.

Thanks again

4

Hi Haddie,

From what we see in the logs it appears that the vesc (the motors for the mobile base wheels) are not detected by the system. I’m saying this because of the line 222

FileNotFoundError: [Errno 2] No such file or directory: ‘/dev/vesc_wheels’

Just to validate that the problem comes from the mobile base and not the rest of the robot, you can edit the config file of the robot and disable the mobile base. The config file is located at ~/.reachy_config/reachy.yaml

The config file should look like either:

serial_number: your_reachy_serial_number

or already have a mobile base entry:

mobile_base:
  enable: true
serial_number: your_reachy_serial_number

To disable the mobile base, just make sure that your config file has the enable key to false for the mobile base:

mobile_base:
  enable: false
serial_number: your_reachy_serial_number

Make sure that you saved the modification, reboot the robot and let me know if you are not still able to connect to the Python sdk. Don’t hesitate to provide logs as well.

Once we have confirmed that the issue comes from the mobile base, we will try to debug it together.

5

Also about the file upload, apparently discourse does not allow to upload .txt file by default but .zip files should be ok, let me know if this works!