kubernetes
/
kops
mirror of https://github.com/kubernetes/kops.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

stop kubelet to prevent orphan containers 

xref #6728

When install the nvidia docker runtime, kubelet will try to restart
stopped containers via old docker. This will leak those containers as
orphan containers.
This commit is contained in:
Hanfei Shen 2019-07-30 21:50:36 +08:00 committed by Hanfei Shen
parent bc6e26ef43
commit 7b8c520e20
No known key found for this signature in database
GPG Key ID: 8645106F29EA1F55
1 changed files with 8 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
hooks/nvidia-device-plugin/image/files/02-nvidia-docker.sh
View File

@ -66,6 +66,11 @@ EOF
# Note that the nvidia-docker version must match the docker-ce version
# --force-confold prevents prompt for replacement of daemon.json
apt-get -y update
# Stop protokube to ensure not bring kubelet up again
systemctl stop protokube
# Stop kubelet to ensure not bring stopped containers up again and leak
# them as orphan containers
systemctl stop kubelet
# pin versions https://github.com/NVIDIA/nvidia-docker/wiki/Frequently-Asked-Questions#how-do-i-install-20-if-im-not-using-the-latest-docker-version
apt-get install -y --allow-downgrades -o Dpkg::Options::="--force-confold" \
nvidia-docker2=2.0.3+docker18.09.4-1 \
@ -79,3 +84,6 @@ apt-get install -y --allow-downgrades -o Dpkg::Options::="--force-confold" \
# Permanently disable these systemd units via masking.
systemctl mask cloud-init.service
systemctl mask kops-configuration.service
# Restore protokube and protokube will bring up kubelet
systemctl start protokube