https://rancher.com/ logo
Join Slack
Powered by
This message was deleted.
# rancher-desktop
a
This message was deleted.
w
is docker/nerdctl working just for a single container instance?
s
Yes the docker example show in the same how-to guide seems to work fine
f
I always start with checking basic k8s functionality first:
Copy code
$ kubectl get no
NAME                   STATUS   ROLES                  AGE   VERSION
lima-rancher-desktop   Ready    control-plane,master   45h   v1.24.6+k3s1
$ kubectl get po -A
NAMESPACE     NAME                                      READY   STATUS      RESTARTS      AGE
kube-system   helm-install-traefik-crd-2j8s4            0/1     Completed   0             45h
kube-system   helm-install-traefik-vxj7p                0/1     Completed   1             45h
kube-system   svclb-traefik-67404bd4-gcbkh              2/2     Running     0             45h
kube-system   traefik-7cd4fcff68-csxsb                  1/1     Running     0             45h
kube-system   local-path-provisioner-7b7dc8d6f5-f8vtm   1/1     Running     1 (49s ago)   45h
kube-system   metrics-server-668d979685-lqnkp           1/1     Running     1 (49s ago)   45h
kube-system   coredns-b96499967-84nbd                   1/1     Running     0             45h
All pods should be running, with the exception of the 
helm-*
ones that are jobs that run and finish.
s
Here I get the same error, infinite lines of errors like 
W1006 18:00:03.676804  79996 transport.go:243] Unable to cancel request for *exec.roundTripper
f
I would go to the Troubleshooting page and run "Reset Kubernetes"
s
I did a “reset kubernetes”, but the command 
kubectl get no
showed the exact same error again…
f
Can you specify more information about your setup: which macOS version, Intel or ARM, which Rancher Desktop version, which Kubernetes version, how much memory and CPU have you allocated for the RD VM?
And then we probably need to see the logs, if the setup otherwise seems fine
s
• MacOS 12.6 • dual core i5 • 8GB memory • rancher desktop version 1.5.1 • kubernetes: n/a (How to find out?) • VM allocation: 2 CPU, 2 GB RAM
kubernetes version 1.24.6
In the 
k3s
log I see an error like 
"Proxy error: write failed: write tcp 127.0.0.1:6443->127.0.0.1:33740: write: connection reset by peer"
f
I don't know, but it may be that the machine is not powerful enough to run Rancher Desktop and Kubernetes. Maybe things get so slow that timeouts prevent kubernetes from properly initializing.
s
Yes that could be well the case. But I am not sure anyway if I require kubernetes. docker seems to work, so at least that is fine…
Thanks for your time and help!
f
If you don't need kubernetes, then you should definitely disable it, to get more memory and CPU resources for docker. Are you using the containerd or moby engine?
s
moby
f
I'm just trying to repro your setup, but the closest machine I have is a dual-core i7 with 16GB of memory
Just starting RD with 2 GB of memory and moby to check, but have a faster machine and more memory might be enough to still make it work (normally I would always give it at least 4GB of RAM on a 16GB machine)
Yeah, it still works fine on that machine with 2GB for the VM.
You could try running Activity Monitor to see if the machine is running at full CPU load. If it is, then that would be the cause of your problem
My machine is still 80% idle with this setup
s
That does not seem to be the case as I am listening to music at the same time. No breakups in the music, so I am probably not at 100%
f
Can you run Activity Monitor, just to see what the usage is?
s
CPU usage seems to be less than 50%
(when not doing anything)
f
Hmmm, then I don't know
s
No worries, thank you anyway
👍 1
f
You could upload the logs, but if you don't really care about kubernetes, then you could also disable it and no longer worry about it
👍 1
1259 Views
Open in Slack
PreviousNext
Powered by