Processes kubectl outputs to enable vim-like navigation in a buffer for your cluster.
Navigate your cluster in a buffer, using hierarchy where possible (backspace for up, enter for down) e.g. root -> deplyoment -> pod -> container
Exec into containers
In the pod view, select a pod by pressing<cr>
and then again <cr>
on the container you want to exec into
Aliases (fallback view)
A fallback view that directs custom resources and has basic functionality such desc, edit, del- kubectl
- curl
- neovim >= 0.10
Install the plugin with your preferred package manager:
return {
{
"ramilito/kubectl.nvim",
config = function()
require("kubectl").setup()
end,
},
}
We expose open, close and toggle to bind against:
vim.keymap.set("n", "<leader>k", '<cmd>lua require("kubectl").toggle()<cr>', { noremap = true, silent = true })
You can override the plugin's keymaps using the <Plug>
mappings:
Default Mappings
-- default mappings
local group = vim.api.nvim_create_augroup("kubectl_mappings", { clear = true })
vim.api.nvim_create_autocmd("FileType", {
group = group,
pattern = "k8s_*",
callback = function(ev)
local k = vim.keymap.set
local opts = { buffer = ev.buf }
-- Global
k("n", "<Plug>(kubectl.help)", "g?", opts) -- Help float
k("n", "<Plug>(kubectl.refresh)", "gr", opts) -- Refresh view
k("n", "<Plug>(kubectl.sort)", "gs", opts) -- Sort by column
k("n", "<Plug>(kubectl.delete)", "gD", opts) -- Delete resource
k("n", "<Plug>(kubectl.describe)", "gd", opts) -- Describe resource
k("n", "<Plug>(kubectl.yaml)", "gy", opts) -- View yaml
k("n", "<Plug>(kubectl.edit)", "ge", opts) -- Edit resource
k("n", "<Plug>(kubectl.filter_label)", "<C-l>", opts) -- Filter labels
k("n", "<Plug>(kubectl.go_up)", "<BS>", opts) -- Go back to previous view
k("v", "<Plug>(kubectl.filter_term)", "<C-f>", opts) -- Filter selected text
k("n", "<Plug>(kubectl.select)", "<CR>", opts) -- Resource select action (different on each view)
k("n", "<Plug>(kubectl.tab)", "<Tab>", opts) -- Tab completion (ascending, when applicable)
k("n", "<Plug>(kubectl.shift_tab)", "<Tab>", opts) -- Tab completion (descending, when applicable)
k("n", "<Plug>(kubectl.quit)", "", opts) -- Close view (when applicable)
k("n", "<Plug>(kubectl.kill)", "gk", opts) -- Pod/portforward kill
k("n", "<Plug>(kubectl.toggle_headers)", "<M-h>", opts) -- Toggle headers
-- Views
k("n", "<Plug>(kubectl.alias_view)", "<C-a>", opts) -- Aliases view
k("n", "<Plug>(kubectl.contexts_view)", "<C-x>", opts) -- Contexts view
k("n", "<Plug>(kubectl.filter_view)", "<C-f>", opts) -- Filter view
k("n", "<Plug>(kubectl.namespace_view)", "<C-n>", opts) -- Namespaces view
k("n", "<Plug>(kubectl.portforwards_view)", "gP", opts) -- Portforwards view
k("n", "<Plug>(kubectl.view_deployments)", "1", opts) -- Deployments view
k("n", "<Plug>(kubectl.view_pods)", "2", opts) -- Pods view
k("n", "<Plug>(kubectl.view_configmaps)", "3", opts) -- ConfigMaps view
k("n", "<Plug>(kubectl.view_secrets)", "4", opts) -- Secrets view
k("n", "<Plug>(kubectl.view_services)", "5", opts) -- Services view
k("n", "<Plug>(kubectl.view_ingresses)", "6", opts) -- Ingresses view
k("n", "<Plug>(kubectl.view_api_resources)", "", opts) -- API-Resources view
k("n", "<Plug>(kubectl.view_clusterrolebinding)", "", opts) -- ClusterRoleBindings view
k("n", "<Plug>(kubectl.view_crds)", "", opts) -- CRDs view
k("n", "<Plug>(kubectl.view_cronjobs)", "", opts) -- CronJobs view
k("n", "<Plug>(kubectl.view_daemonsets)", "", opts) -- DaemonSets view
k("n", "<Plug>(kubectl.view_events)", "", opts) -- Events view
k("n", "<Plug>(kubectl.view_helm)", "", opts) -- Helm view
k("n", "<Plug>(kubectl.view_jobs)", "", opts) -- Jobs view
k("n", "<Plug>(kubectl.view_nodes)", "", opts) -- Nodes view
k("n", "<Plug>(kubectl.view_overview)", "", opts) -- Overview view
k("n", "<Plug>(kubectl.view_pv)", "", opts) -- PersistentVolumes view
k("n", "<Plug>(kubectl.view_pvc)", "", opts) -- PersistentVolumeClaims view
k("n", "<Plug>(kubectl.view_sa)", "", opts) -- ServiceAccounts view
k("n", "<Plug>(kubectl.view_top_nodes)", "", opts) -- Top view for nodes
k("n", "<Plug>(kubectl.view_top_pods)", "", opts) -- Top view for pods
-- Deployment/DaemonSet actions
k("n", "<Plug>(kubectl.rollout_restart)", "grr", opts) -- Rollout restart
k("n", "<Plug>(kubectl.scale)", "gss", opts) -- Scale workload
k("n", "<Plug>(kubectl.set_image)", "gi", opts) -- Set image (only if 1 container)
-- Pod/Container logs
k("n", "<Plug>(kubectl.logs)", "gl", opts) -- Logs view
k("n", "<Plug>(kubectl.history)", "gh", opts) -- Change logs --since= flag
k("n", "<Plug>(kubectl.follow)", "f", opts) -- Follow logs
k("n", "<Plug>(kubectl.wrap)", "gw", opts) -- Toggle wrap log lines
k("n", "<Plug>(kubectl.prefix)", "gp", opts) -- Toggle container name prefix
k("n", "<Plug>(kubectl.timestamps)", "gt", opts) -- Toggle timestamps prefix
k("n", "<Plug>(kubectl.previous_logs)", "gpp", opts) -- Toggle show previous logs
-- Node actions
k("n", "<Plug>(kubectl.cordon)", "gC", opts) -- Cordon node
k("n", "<Plug>(kubectl.uncordon)", "gU", opts) -- Uncordon node
k("n", "<Plug>(kubectl.drain)", "gR", opts) -- Drain node
-- Top actions
k("n", "<Plug>(kubectl.top_nodes)", "gn", opts) -- Top nodes
k("n", "<Plug>(kubectl.top_pods)", "gp", opts) -- Top pods
-- CronJob actions
k("n", "<Plug>(kubectl.suspend_cronjob)", "gss", opts) -- Suspend CronJob
k("n", "<Plug>(kubectl.create_job)", "gc", opts) -- Create Job from CronJob
k("n", "<Plug>(kubectl.portforward)", "gp", opts) -- Pods/Services portforward
k("n", "<Plug>(kubectl.browse)", "gx", opts) -- Ingress view
k("n", "<Plug>(kubectl.yaml)", "gy", opts) -- Helm view
end,
})
For overriding the default mappings when using lazy.nvim
check out our wiki page.
{
log_level = vim.log.levels.INFO,
auto_refresh = {
enabled = true,
interval = 300, -- milliseconds
},
diff = {
bin = "kubediff" -- or any other binary
},
kubectl_cmd = { cmd = "kubectl", env = {}, args = {} },
namespace = "All",
namespace_fallback = {}, -- If you have limited access you can list all the namespaces here
hints = true,
context = true,
heartbeat = true,
kubernetes_versions = true,
lineage = {
enabled = false, -- This feature is in beta at the moment
},
logs = {
prefix = true,
timestamps = true,
since = "5m"
},
alias = {
apply_on_select_from_history = true,
max_history = 5,
},
filter = {
apply_on_select_from_history = true,
max_history = 10,
},
float_size = {
-- Almost fullscreen:
-- width = 1.0,
-- height = 0.95, -- Setting it to 1 will cause bottom to be cutoff by statuscolumn
-- For more context aware size:
width = 0.9,
height = 0.8,
-- Might need to tweak these to get it centered when float is smaller
col = 10,
row = 5,
},
obj_fresh = 5, -- highlight if creation newer than number (in minutes)
}
The plugin uses the following highlight groups:
Highlight Groups
The setup function only adds ~1ms to startup. We use kubectl proxy and curl to reduce latency.
We leverage the Kubernetes Informer to efficiently monitor resource updates.
By using the resourceVersion
, we avoid fetching all resources in each loop.
Instead, the Informer provides only the changes, significantly reducing overhead and improving performance.
As we advance to v1.0.0
, our primary goal is to maintain the stability of the
plugin and minimize any breaking changes. We are committed to providing a
reliable and consistent user experience.
This plugins main purpose is to browse the kubernetes state using vim like navigation and keys, similar to oil.nvim for file browsing.