You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have tried different drivers (docker, kvm2, qemu) and different CNIs (bridge, kindnet, calico), but I can't seem to get the gvisor addon to work. Using the docker driver, I get permission denied errors (see https://stackoverflow.com/questions/72555998/cant-launch-pod-using-gvisor). I tried switching CNIs without success.
So I tried switching to a VM based solution (kvm2/qemu) but now once I enable the gVisor addon, the system goes into a crash loop (similar to #5740). The pod description gives the same error on each pod:
Normal SandboxChanged 5m19s (x4 over 7m26s) kubelet Pod sandbox changed, it will be killed and re-created.
Warning BackOff 2m19s (x21 over 6m51s) kubelet Back-off restarting failed container
$ kubectl get pod,runtimeclass gvisor -n kube-system
NAME READY STATUS RESTARTS AGE
pod/gvisor 0/1 CrashLoopBackOff 6 (3m15s ago) 11m
NAME HANDLER AGE
runtimeclass.node.k8s.io/gvisor runsc 11m
runsc version
runsc version release-20200211.0-16-gb2e86906ea4f
spec: 1.0.1-dev
containerd has updated its configuration format (once again). Minikube breaks because the plugin is trying to use the old format. Let me do a quick fix for now...we'll need a better way to patch config.toml to configure the runtime. Right now, it replaces the entire file and may lose other configuration changes.
Description
I have tried different drivers (docker, kvm2, qemu) and different CNIs (bridge, kindnet, calico), but I can't seem to get the gvisor addon to work. Using the docker driver, I get
permission denied
errors (see https://stackoverflow.com/questions/72555998/cant-launch-pod-using-gvisor). I tried switching CNIs without success.So I tried switching to a VM based solution (kvm2/qemu) but now once I enable the gVisor addon, the system goes into a crash loop (similar to #5740). The pod description gives the same error on each pod:
Steps to reproduce
I followed the instructions here: https://github.com/kubernetes/minikube/blob/master/deploy/addons/gvisor/README.md
runsc version
docker version (if using docker)
# crictl version Version: 0.1.0 RuntimeName: containerd RuntimeVersion: 1.6.6 RuntimeApiVersion: v1alpha2
uname
Linux euclid 5.15.59-1-MANJARO #1 SMP PREEMPT Wed Aug 3 11:20:04 UTC 2022 x86_64 GNU/Linux
kubectl (if using Kubernetes)
repo state (if built from source)
No response
runsc debug logs (if available)
The text was updated successfully, but these errors were encountered: