I'm not sure if it was because of the context matching up or if it was coincidence, but I wanted to note it here in case that helps anyone. Once I had switched to that context, I noticed that kubernetes started. I switched to that context (my context was on another k8s cluster for work).
My ~/.docker/config.json had a context for the docker desktop kubernetes. Restarted my machine, but k8s didn't appear to start or was taking a while and I got impatient to try the next thing. I also found an extra mapping to 127.0.0.1 auto-generated by Docker Desktop (perhaps an earlier version?) and deleted that. Updated /etc/hosts/ as per abdennour's comment. 
Tried rm -rf ~/Library/Group\ Containers//pki/ as per gstruczynski's comment. I'll list everything I tried in case this helps people browsing through google, or in case my extra tweaks were the fix. I had this issue and after trying a few advices in this thread and a few of my own fiddlings, I fixed it. After services running (Docker and Kubernetes), try access the API: 20:23:55.565458-0300 localhost Docker: (CoreFoundation) Bundle:, key: kubeStateRunning, value:, table: Localizable, localizationName: (null), result: Kubernetes is runningĦ. 20:23:55.563460-0300 localhost Docker: (CoreFoundation) Bundle:, key: kubeStateRunning, value:, table: Localizable, localizationName: (null), result: Kubernetes is running After some time (and a lot of informations), tailing will show following messages:. (It takes about 1 minute 30 seconds to startup) Reinstall Kubernetes from Docker Preferences Panel (MUST contain docker-for-desktop name on first localhost line)Ĥ. kube file (rm -rf ~/.kube) - if presentģ. Uninstalled Kubernetes from Docker Preferences Panel (if present)Ģ. $ /usr/bin/log stream -style syslog -level=debug -color=always -predicate "$pred"ġ. $ pred='process matches ".*(ocker|vpnkit).*" || (process in & eventMessage contains "docker")' 
Open terminal and follow up the logs of Docker for Desktop:
0 kommentar(er)
