Open wangjiangyong opened 2 years ago
Are you running as root
? This looks like NFS root squashing. I suppose you can't do ls /home/wjy/nv-maskrcnn/dataset_dir/coco2017
as root
either.
yes,root user. change to root directory, there is no permission problem. But why is there permission problem, root is the highest privilege
[root@node0 pytorch]# srun --mpi=none --container-image=./mlperf-nvidia+object_detection.sqsh --container-name=object_detection --container-mounts=/home/wjy/nv-maskrcnn/dataset_dir/coco2017:/data ls /data slurmstepd: error: pyxis: container start failed with error code: 1 slurmstepd: error: pyxis: printing enroot log file: slurmstepd: error: pyxis: enroot-mount: failed to mount: /home/wjy/nv-maskrcnn/dataset_dir/coco2017 at /tmp/enroot-data/user-0/pyxis_object_detection/data: Permission denied slurmstepd: error: pyxis: couldn't start container slurmstepd: error: spank: required plugin spank_pyxis.so: task_init() failed with rc=-1 slurmstepd: error: Failed to invoke spank plugin stack srun: error: node0: task 0: Exited with exit code 1