Host OS: Ubuntu 20.04 Kubernetes: 1.23 RAM: 64g HDD: 3.7Tb

Containers are failing and rescheduling due to low ephemeral storage. Ex. The node was low on resource: ephemeral-storage. Container <container-name> was using 24856Ki, which exceeds its request of 0.

I have already configured the docker daemon.json to limit log sizes, which seemed to work, but now that I've added more containers I am running out of memory again.

I am wondering what best practice is here for managing this storage. The behavior I desire would be for the pods to just keep running and dump excess logs as needed without having to restart. However, it is not clear to me if this is possible.

Can anyone recommend a strategy here?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.