The pod deployment is missing a node selector for linux nodes, so in a hybrid cluster you can end up with sshjump failing to start because its scheduled to a windows node.
Fix would be to update the 'kubectl run' command in the plugin to deploy a manifest with a node selector.
@swgriffith PR #9 has been merged! I've release 0.5.0 which can be installed via kubectl krew install ssh-jump
I'm closing the issue
Thank you so much for your contribution again!
The pod deployment is missing a node selector for linux nodes, so in a hybrid cluster you can end up with sshjump failing to start because its scheduled to a windows node.
Fix would be to update the 'kubectl run' command in the plugin to deploy a manifest with a node selector.