Skip to content

slurm: set containerd root to EBS #914

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
maekawataiki wants to merge 3 commits into
base: main
Choose a base branch
from
Open

slurm: set containerd root to EBS #914

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f54afb2
fix containerd directory
maekawataiki Dec 7, 2025
e4583d6
Update Docker and containerd configuration in script
KeitaW Dec 17, 2025
d1b87e3
Remove state configuration from install_docker.sh
KeitaW Dec 17, 2025
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
21 changes: 20 additions & 1 deletion 1.architectures/5.sagemaker-hyperpod/LifecycleScripts/base-config/utils/install_docker.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,6 +56,8 @@ sudo usermod -aG docker ubuntu
# See: https://github.com/aws-samples/awsome-distributed-training/issues/127
#
# Docker workdir doesn't like Lustre. Tried with storage driver overlay2, fuse-overlayfs, & vfs.
# Also, containerd ships with a commented root in its default config; we need to ensure an
# uncommented root that points to the fast local volume.
if [[ $(mount | grep /opt/sagemaker) ]]; then
cat <<EOL >> /etc/docker/daemon.json
{
Expand All @@ -66,6 +68,14 @@ EOL
sed -i \
's|^\[Service\]$|[Service]\nEnvironment="DOCKER_TMPDIR=/opt/sagemaker/docker/tmp"|' \
/usr/lib/systemd/system/docker.service

# Ensure containerd config exists and point its root to /opt/sagemaker
if [[ ! -f /etc/containerd/config.toml ]]; then
containerd config default | sudo tee /etc/containerd/config.toml >/dev/null
fi
sudo sed -i \
-e 's|^#\\?root *=.*|root = "/opt/sagemaker/docker/containerd"|' \
Copy link

@PremiumSpider PremiumSpider Jan 16, 2026

Choose a reason for hiding this comment

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

Tested these changes but it didn't work but the below did
sudo sed -i -e 's|^#\?root *=.*|root = "/opt/sagemaker/docker/containerd"|'

Copy link
Collaborator

@shimomut shimomut Jan 21, 2026

Choose a reason for hiding this comment

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

Could you use /opt/sagemaker/containerd/data-root instead of /opt/sagemaker/docker/containerd? for consistency with HyperPod EKS side.

See: https://github.com/aws-samples/awsome-distributed-training/blob/main/1.architectures/7.sagemaker-hyperpod-eks/LifecycleScripts/base-config/on_create_main.sh#L70

/etc/containerd/config.toml
elif [[ $(mount | grep /opt/dlami/nvme) ]]; then
cat <<EOL >> /etc/docker/daemon.json
{
Expand All @@ -76,7 +86,16 @@ EOL
sed -i \
's|^\[Service\]$|[Service]\nEnvironment="DOCKER_TMPDIR=/opt/dlami/nvme/docker/tmp"|' \
/usr/lib/systemd/system/docker.service

# Ensure containerd config exists and point its root to /opt/dlami/nvme
if [[ ! -f /etc/containerd/config.toml ]]; then
containerd config default | sudo tee /etc/containerd/config.toml >/dev/null
fi
sudo sed -i \
-e 's|^#\\?root *=.*|root = "/opt/dlami/nvme/docker/containerd"|' \
Copy link

@PremiumSpider PremiumSpider Jan 16, 2026

Choose a reason for hiding this comment

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

Tested these changes but it didn't work but the below did
sudo sed -i -e 's|^#\?root *=.*|root = "/opt/sagemaker/docker/containerd"|'

/etc/containerd/config.toml
fi

systemctl daemon-reload
systemctl restart docker
systemctl restart containerd
systemctl restart docker