View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0014172||CentOS-7||-OTHER||public||2017-11-23 22:00||2017-11-23 22:00|
|Platform||Amazon AWS - Kubernetes||OS||CentOS7||OS Version|
|Target Version||Fixed in Version|
|Summary||0014172: oom crashes the machine|
|Description||I am running a Kubernetes cluster composed of CentOS-7 machines. I detected that whenever the cgroup where the kubernetes pods & kubelet cannot allocate more memory. The machine crashes. It seems to be related to an oom error that crashes the Kernel.|
I tried to set `vm.overcommit_memory=1` or to `2` but I get the same result. I followed what suggested in this page https://docs.openshift.org/latest/admin_guide/overcommit.html. But I believe there should be something else to configure when using CentOS.
Primarily I'd like to avoid my machine to crash due to this oom problem. How can I achieve this contention ?
PRETTY_NAME="CentOS Linux 7 (Core)"
|Steps To Reproduce||I create process inside one cgroup allocating more memory that the one available for this cgroup.|
|Additional Information||This issue could be related to https://lists.centos.org/pipermail/centos/2012-May/126089.html. I also tried what it was mentioned in this issue without any success.|
|Tags||No tags attached.|