Volumes disconnect when using the aws-efs-csi-driver

Has anyone tried using the aws-efs-csi-driver with any long term success? I’m running a single node rancher (v2.7.5) install using docker to test this out and I can mount the EFS as a volume in a workload but after a certain amount of time (4 to 48 hours), the mount goes into an unusable state.

This is the repo I’m pointing to and installing to my cluster using the provided helm chart.