Expand description
cgroup — Namespace and cgroup resource-isolation primitives.
This slice’s concrete target: resolving the real cgroup path of an
already-running process on the host, by pid — filling
principal::ContainerRef::cgroup, which is unconditionally None until
something actually resolves it. lifecycle::Lifecycle::container_ref
is the caller that wires this in for a real dockerd container.
Scope boundary (hard constraint, same as every substrate-kit
module): this is OS mechanics only. It resolves what a cgroup path
is, never why a cgroup limit is set to a given value or who should
be allowed to see/change it — that’s an authorization/policy question
for identity-kit to answer, not this crate. See principal.rs’s
module doc for the full hard-constraint statement.
Known limitation, not an oversight: this resolver only understands
the cgroup v2 unified hierarchy (/proc/<pid>/cgroup’s 0::<path>
line), matching the target hosts (dockerd with the systemd
cgroup driver defaults to cgroup v2 on any reasonably current kernel).
A cgroup v1 or hybrid host — no 0:: line present — resolves to
SubstrateError::Resolve rather than guessing at a v1 controller
hierarchy; broader support is real follow-up work if a v1 host ever
needs to run this substrate.
Full namespace/cgroup isolation primitives (creating/limiting a
cgroup, not just reading one back) remain deferred; resource limits on a
container are not yet part of lifecycle::ContainerSpec either.
Constants§
- CGROUP_
V2_ MOUNT - The cgroup v2 unified hierarchy’s standard mount point. Universal on every distro this substrate targets — not configurable here because none of its target hosts mounts cgroup v2 anywhere else.
Functions§
- resolve_
cgroup_ path - Resolve the real, absolute cgroup v2 path for a running process on the host, by pid.