Unable to run pod in Kubernetes

[gcpguru]

When the pod starts up, it immediately fails with the following:

Traceback (most recent call last):
File "/usr/local/bin/ncs2_device_plugin", line 11, in
load_entry_point('ncs2-device-plugin==0.2.0', 'console_scripts', 'ncs2_device_plugin')()
File "/app/ncs2_device_plugin/main.py", line 46, in main
server = NCS2DevicePluginServer(service=service, logger=logger)
File "/app/ncs2_device_plugin/server.py", line 25, in init
self.server.add_insecure_port('unix://' + ncs2_socket_path)
File "/usr/local/lib/python3.8/dist-packages/grpcio-1.34.0-py3.8-linux-x86_64.egg/grpc/_server.py", line 961, in add_insecure_port
return _common.validate_port_binding_result(
File "/usr/local/lib/python3.8/dist-packages/grpcio-1.34.0-py3.8-linux-x86_64.egg/grpc/_common.py", line 166, in validate_port_binding_result
raise RuntimeError(_ERROR_MESSAGE_PORT_BINDING_FAILED % address)
RuntimeError: Failed to bind to address unix:///var/lib/kubelet/device-plugins/ncs2.sock

Complete log file attached.
failed_pod.log

[pmundt]

Thanks for the report @rafaeljcuevas. Can you try re-run this with GRPC_VERBOSITY=debug set so we can figure out why the socket file is not being created?

Also, are you using a specific Kubernetes version or distribution?