Running Isaac from standalone scripts seems to ignore persistence container cache

As the title describes, each time I restart a container and run the Isaac from standalone scripts, it ignores the persistent cache and takes ages to open it. This is not the case if I run Isaac with runheadless- bash scripts. Is there a setting I need to make to use the existing cache?

1 Like

I am facing the same issue for more than a year across different versions.

We’ve updated the docker run command for the latest Isaac Sim 2022.2.1 container. These mounts below should help get persistent cache on host:

-v ~/docker/isaac-sim/kit/cache/Kit:/isaac-sim/kit/cache/Kit:rw \
-v ~/docker/isaac-sim/cache/ov:/root/.cache/ov:rw \
-v ~/docker/isaac-sim/cache/pip:/root/.cache/pip:rw \
-v ~/docker/isaac-sim/cache/glcache:/root/.cache/nvidia/GLCache:rw \
-v ~/docker/isaac-sim/cache/computecache:/root/.nv/ComputeCache:rw \

Thanks!

Should the first line be as you wrote below

or

-v ~/docker/isaac-sim/cache/kit:/isaac-sim/kit/cache/Kit:rw \

as in (https://github.com/NVIDIA-Omniverse/OmniIsaacGymEnvs/blob/main/docker/run_docker.sh)?

Thanks for catching that. It’s a typo. It should be:

-v ~/docker/isaac-sim/cache/kit:/isaac-sim/kit/cache/Kit:rw \

However, it should really matter since you can use any path you want :)

For Isaac Sim 2022.2.1, you could remove these lines too:

-v /etc/vulkan/icd.d/nvidia_icd.json:/etc/vulkan/icd.d/nvidia_icd.json
-v /etc/vulkan/implicit_layer.d/nvidia_layers.json:/etc/vulkan/implicit_layer.d/nvidia_layers.json \

Thanks. this seems to solve the issue. The consequent runs of the Isaac Sim from standalone scripts use the cache.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.