crictl within VM fails with "connection is unavailable" due to missing /etc/crictl.yaml

[afbjorklund]

Is this a BUG REPORT or FEATURE REQUEST? (choose one):

FEATURE REQUEST

** Details**

When configuring CRI-O as the container runtime, we should configure crictl to use it by default.

$ cat /etc/crictl.yaml
runtime-endpoint: unix:///var/run/dockershim.sock
image-endpoint: unix:///var/run/dockershim.sock

$ cat /etc/crictl.yaml
runtime-endpoint: unix:///var/run/crio/crio.sock
image-endpoint: unix:///var/run/crio/crio.sock

Otherwise using crictl will give an error, or maybe some ugly looking deprecation warning etc.

$ sudo crictl version
FATA[0000] getting the runtime version failed: rpc error: code = Unavailable desc = grpc: the connection is unavailable 
$ sudo crictl -r /var/run/crio/crio.sock version
W0803 12:03:56.591607   15155 util_unix.go:75] Using "/var/run/crio/crio.sock" as endpoint is deprecated, please consider using full url format "unix:///var/run/crio/crio.sock".
Version:  0.1.0
RuntimeName:  cri-o
RuntimeVersion:  1.10.0
RuntimeApiVersion:  v1alpha1

[afbjorklund]

Related to #3024 ($CONTAINER_RUNTIME_ENDPOINT)

[tstromberg]

Confirmed. When this is used:

minikube start \
         --network-plugin=cni \
         --container-runtime=cri-o \
         --bootstrapper=kubeadm --vm-driver=kvm2

crictl version fails within "minikube ssh":

FATA[0000] getting the runtime version failed: rpc error: code = Unavailable desc = grpc: the connection is unavailable 

This fixes it:

cat > /etc/crictl.yaml
runtime-endpoint: unix:///var/run/crio/crio.sock
image-endpoint: unix:///var/run/crio/crio.sock

[afbjorklund]

Sadly this needs to be set up by the provisioner, which I am not even sure knows the container-runtime...

https://github.com/kubernetes/minikube/blob/master/pkg/provision/buildroot.go#L188

[afbjorklund]

No, the provisioner indeed only knows about Docker - and only the subset config hardcoded by machine.

https://github.com/kubernetes/minikube/blob/master/vendor/github.com/docker/machine/libmachine/engine/engine.go#L7:L21

This means that setting up registries and setting up the crictl configuration needs to be in bootstrapper.

https://github.com/kubernetes/minikube/blob/master/pkg/minikube/bootstrapper/kubeadm/kubeadm.go#L208

Which is probably wrong, since crictl doesn't have anything to do with kubeadm (it's a requirement).

But we can't install it with crictl, since we don't know which of the three endpoints will ultimately be used.

[afbjorklund]

@tstromberg : should be fixed now...

Note that you will still get the error above, unless you remember to run: sudo crictl version

                         _             _            
            _         _ ( )           ( )           
  ___ ___  (_)  ___  (_)| |/')  _   _ | |_      __  
/' _ ` _ `\| |/' _ `\| || , <  ( ) ( )| '_`\  /'__`\
| ( ) ( ) || || ( ) || || |\`\ | (_) || |_) )(  ___/
(_) (_) (_)(_)(_) (_)(_)(_) (_)`\___/'(_,__/'`\____)

$ crictl version
FATA[0000] getting the runtime version failed: rpc error: code = Unavailable desc = grpc: the connection is unavailable 
$ sudo crictl version
Version:  0.1.0
RuntimeName:  cri-o
RuntimeVersion:  1.10.0
RuntimeApiVersion:  v1alpha1
$ 

At least it doesn't segfault anymore, when running crictl version ? Always something...

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale