[Bug Report] CrashLoopBackOff in `Metadatacontroller-0` pod

[RyanKor]

1. Issue & Problem

• Metadatacontroller-0 does not become "RUNNING" status after it enters CrashLoopBackOff
• kubectl log returns the message below.

kubelet Error: failed to start container "metacontroller": Error response from daemon: OCI runtime create failed: container_linux.go:380: starting container process caused: process_linux.go:545: container init caused: process_linux.go:508: setting cgroup config for procHooks process caused: failed to write "400000": write /sys/fs/cgroup/cpu/kubepods/burstable/pod8dccff53-140f-4479-9442-49887bc6bc2c/metacontroller/cpu.cfs_quota_us: invalid argument: unknown

2. Solution

• The limited computing resoures result in the problem.
• You should expand your resources at least the spec below

CPU : 6 core RAM : 12GB DISK : 50GB

• You can use cloud services like GCP, AWS, Azure or allocate your local's ones using the command like this. (If you are using minikube)
• > minikube start --cpus 6 --memory 12288 --disk-size=50g
• Remember Kubeflow does not offer a command to change your resources.