Fix detection of docker cgroup on RHEL #25907
Conversation
Check docker's pid file, then fallback to pidof when trying to determine the pid for docker. The latest docker RPM for RHEL changes /usr/bin/docker from an executable to a shell script (to support /usr/bin/docker-current and /usr/bin/docker-latest). The pidof check for docker fails in this case, so we check /var/run/docker.pid first (the default location), and fallback to pidof if that fails.
ncdc
added
sig/node
k8s-github-robot
added
the
size/M
|
LGTM. /cc @vishh - we need to ensure that the unit for otherwise, runtime stats are not accurate for docker runtime (it will default to / as the since that will be the cpu cgroup in /proc/../cgroup file) when accounting is off. We really should ensure that cpu and memory accounting is enabled in a follow-on and there cgroup name is the same... |
|
For reference: |
|
LGTM Thanks for fixing this. |
dchen1107
added
the
lgtm
|
LGTM. @derekwaynecarr If metrics are not available, we will be returning |
|
Today it returns runtime stats for root cgroup. Confirmed today. On Thursday, May 19, 2016, Kubernetes Bot notifications@github.com wrote:
|
|
@k8s-bot test this [submit-queue is verifying that this PR is safe to merge] |
|
GCE e2e build/test passed for commit 6744a74. |
|
Automatic merge from submit-queue |
Check docker's pid file, then fallback to pidof when trying to determine the pid for docker. The
latest docker RPM for RHEL changes /usr/bin/docker from an executable to a shell script (to support
/usr/bin/docker-current and /usr/bin/docker-latest). The pidof check for docker fails in this case,
so we check /var/run/docker.pid first (the default location), and fallback to pidof if that fails.
@kubernetes/sig-node @kubernetes/rh-cluster-infra