Replies: 2 comments 8 replies
-
Because there is no easy way to delegate cgroup v2 without systemd |
Beta Was this translation helpful? Give feedback.
-
|
I'm a bit confused, when running dind-rootless as the root user, it supports the cgroupfs driver. What is the reason for the difference in support for the cgroupfs driver between the root and non-root users inside the container？
|
Beta Was this translation helpful? Give feedback.
-
Just a beautiful mistask，but it works well 😁 |
Beta Was this translation helpful? Give feedback.
-
Yes, it should be easy to implement it.
Yes, I have completely forgotten that this was already implemented three years ago 😅 |
Beta Was this translation helpful? Give feedback.
-
Great! In which version Docker will implement this feature? |
Beta Was this translation helpful? Give feedback.
-
I would like to learn more about which module in cgroupv2 chown cgroup to rootless user, and specifically where is implemented in the Docker code. Additionally, if cgroupv2 can chown cgroup to rootless user, can the same be achieved with cgroupv1 ? |
Beta Was this translation helpful? Give feedback.
-
The first major version right after somebody submits a PR
Chowned by runc (see the link above)
Not considered to be safe |
Beta Was this translation helpful? Give feedback.
-
Rootless-mode cgroup driver only supports systemd, why doesn't it support cgroupfs?
The docker info output information(in contianer dind-rootless) is as follows:
host os info：
Beta Was this translation helpful? Give feedback.
All reactions