Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions manifests/rhoai/runtimes/torch_cuda_241.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@ spec:
spec:
template:
spec:
securityContext:
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should try to avoid changing the security context.

My understanding is that it can be avoided by mounting an empty / tmp dir as workspace or changing the workspace directory

11:48
My understanding is that it can be avoided by mounting an empty / tmp dir as workspace or changing the workspace directory

Copy link
Copy Markdown
Member Author

@abhijeet-dhumal abhijeet-dhumal Nov 19, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ah I see, so user should always opt for the directories like a /tmp dir always writable by any user ?

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please correct me here but, the purpose of podTemplateOverrides with volume mounts was to allow users to mount PVCs at any configurable path they choose (like /workspace, /data, /models, etc.)
But without securityContext specified, User will always get permission denied error because:

  • when user mount a PVC at an arbitrary path, container will run as a user 1001 (or whatever the image defaults to)
  • PVC is usually owned by root (UID 0) or some other UID

runAsUser: 1000
fsGroup: 1000
containers:
- name: node
image: quay.io/modh/training:py311-cuda121-torch241
3 changes: 3 additions & 0 deletions manifests/rhoai/runtimes/torch_cuda_251.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@ spec:
spec:
template:
spec:
securityContext:
runAsUser: 1000
fsGroup: 1000
containers:
- name: node
image: quay.io/modh/training:py311-cuda124-torch251
3 changes: 3 additions & 0 deletions manifests/rhoai/runtimes/torch_rocm_241.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@ spec:
spec:
template:
spec:
securityContext:
runAsUser: 1000
fsGroup: 1000
containers:
- name: node
image: quay.io/modh/training:py311-rocm62-torch241
3 changes: 3 additions & 0 deletions manifests/rhoai/runtimes/torch_rocm_251.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,9 @@ spec:
spec:
template:
spec:
securityContext:
runAsUser: 1000
fsGroup: 1000
containers:
- name: node
image: quay.io/modh/training:py311-rocm62-torch251
Loading