Welcome to the Linux Foundation Forum!

Lab 11 - Containers from the same POD on different nodes???

segev Posts: 5
edited February 2019 in LFS258 Class Forum

Steps 9 and 10 in the lab (create the VIP POD with no 'nodeSelector') are as follows:
9 - Create the pod again. Containers should now be spawning on both nodes. You may see pods for the daemonsets as
10 - Determine where the new containers have been deployed. They should be more evenly spread this time.

My understanding is that the Kubernetes scheduler schedules PODs, once the POD has been assigned to a node the containers within the POD will be created on the same assigned node.

Steps 9 and 10 in the Lab 11 appear to suggest that containers defined in the same POD (the VIP POD) will be distributed across multiple nodes.
To me this is wrong, am I missing something here?


  • Hi @segev, thanks for pointing this out. The verbiage should be revised in a future course update release.
    You are correct, a Pod once scheduled on a Node, all of the Containers from this Pod will run on that same Node. The intent here was to say that without a nodeSelector, the Pods will be distributed more evenly between Nodes.

  • segev
    segev Posts: 5

    Hi Chris, thanks for the reply. I would suggest for future course update on this section, to replace the usage of 'docker ps' commands to track the location of the created containers with 'kubectl get po -o wide' which will show on which node the pods were scheduled. This will move the focus from containers to K8S pods, is more direct method and is easier to use.


  • serewicz
    serewicz Posts: 1,000


    Thank you for the suggestion. The update went out yesterday. But I will keep your comment in mind for next edit, which should be after 1.14 drops, in about a month.



Upcoming Training