WARNING: The object store is using /tmp instead of /dev/shm because /dev/shm has only 67104768 bytes available

Ray Libraries (Data, Train, Tune, Serve) Ray Serve
1

Hi, I use the ray inside my kubeflow VM (basically is the docker container vs code server). When I run the ray serving. There is warning as follow

2022-09-23 07:33:25,887 WARNING services.py:1882 -- WARNING: The object store is using /tmp instead of /dev/shm because /dev/shm has only 67104768 bytes available. This will harm performance! You may be able to free up space by deleting files in /dev/shm. If you are inside a Docker container, you can increase /dev/shm size by passing '--shm-size=2.58gb' to 'docker run' (or add it to the run_options list in a Ray cluster config). Make sure to set this to more than 30% of available RAM.

How I configure this inside the docker container?

2

Are you using Kuberay? The operator should be automatically configuring /dev/shm via a volume mount: KubeRay Operator - KubeRay Docs

If not, I think you can configure the /dev/shm mount as follows: Raising shared memory limit of a Kubernetes container | by Anuj Arora | Dive into ML/AI | Medium

With plain docker containers, you can also use the --shm-size flag.