#CGGQEHPPW Hey folks, I’m running a K3s cluster on an on-prem Ubuntu server (28 vCPU, 64GB RAM, 2TB disk). Recently, /var/lib/rancher/k3s is filling up very fast (currently ~100GB used), especially the storage and agent/containerd directories. I’ve checked the following: • /var/lib/rancher/k3s/storage/ has large PVCs (OpenObserve ~44GB, Redis ~15GB) • /var/lib/rancher/k3s/agent/containerd/io.containerd.snapshotter.v1.overlayfs/snapshots/ has many large snapshot dirs (~2GB each) • Image GC is frequently failing (FreeDiskSpaceFailed, ImageGCFailed events) • I’ve already set kubelet args for aggressive image GC and eviction thresholds Despite that, disk usage keeps climbing. Any insights into what might be causing this rapid growth or how to better manage storage in K3s with containerd and long-running workloads? Should I be handling volume cleanup differently?

Seems like you figured out what’s consuming space and you are unable to reduce any of that?