Setup RKE and rancher from existing cluster after RANCHER VM Crash

I got a kubernetes cluster built on RKE with rancherOS as OS, my cluster is already in production as i explain here (waiting for someone who already debug this kind of error on RancherOs)
So my rancherOS VM cannot start anymore and i cannot access to any config file of the existing cluster or to kubectl command that was on that rancherOs VM.And kubectl was not installed on any nodes. Now i cannot interact with cluster through kubectl or RancherUI. My only way to see cluster ressource is through Kubernetes UI that was deployed as an ingress.

Is it possible recreate a new VM, install rancher and Rke and “import” all config files for that existing cluster? Or any other solution do not destroy my cluster? My cluster is in production, has many deployment so i cannot just start another cluster from scratch.
RancherOS 1.5.3
Rancher 2.3.1
RKE 0.3.1
Kubernetes v1.15.5