Docker – CreateContainerError: Context Deadline Exceeded Solution

dockergitlabhelmkubernetesrancher-2

For a project, I must use big containers (500Mb to 60Gb).

I don't have a precise measurement but when I run containers larger than 3-5Gb with gitlab-runner, I get an error in rancher: CreateContainerError: context deadline exceeded

Our kubernetes cluster is build with rke with rancher as web ui and located in our datacenter.

The bug only appears with gitlab-runner, if I launch docker run ... on the kubernetes node, everything works fine.

Maybe there is a timeout somewhere?

Have any of you ever had this problem?

Thank you.

Best Answer

Ok, it seems that adding these lines in my rke config (cluster.yml) resolve this issue:

services:
  kubelet:
    extra_args:
      runtime-request-timeout: '1h'

Related Solutions

Linux – Understanding Java memory behavior in Docker

When you specify

CMD java -jar /data/uberjar.jar -Xmx128m -server

then the values that you intend to be JVM arguments (-Xmx128m -server) are passed as command line arguments to the Main class in the .jar file. They're available as the args in your public static void main(String... args) method.

Note that this is also true if you're running a main class by name, rather than specifying an executable jar file.

To pass them as JVM arguments, rather than arguments to your program, you need to specify them before the -jar arg.

See https://stackoverflow.com/questions/5536476/passing-arguments-to-jar-which-is-required-by-java-interpreter

Can a Kubernetes pod span nodes

The very idea of pods is that of co-location, that is, being able to define compute-locality where it makes sense (or where it's needed). So, no, per definition, since all containers in a pod will be scheduled on the same node, a pod can not span nodes.

See also my critique of pods here. Having said that, we've introduced the pods concept in DC/OS now with v1.9 as well. There seems to be a certain demand, but my hunch (since I don't have data on it) is that it's mostly a sort of safety blanket for cases where one doesn't or can't go all in, cloud-native wise.

Best Answer

Related Solutions

Linux – Understanding Java memory behavior in Docker

Can a Kubernetes pod span nodes

Related Topic