KubeVirt Tools

Phoenix web app for cluster-wide KubeVirt and Kubernetes visibility: sign in with a kubeconfig, then explore a LiveView dashboard backed by a one-shot API snapshot (refresh on demand).

What you get

• Dashboard — VM / VMI counts, node capacity charts, PVC breakdowns, optional Prometheus overlays for cluster usage and node CPU/memory when metrics-server is missing.
• VMs — VirtualMachine list with VMI join and orphan VMI section.
• Networks — Template interfaces vs live VMI network data.
• Disks — Volume and PVC-backed disk detail.
• Storage classes — StorageClass inventory with provisioner, reclaim policy, binding mode, expansion, parameters, and PVC counts (plus warnings for mismatched PVC references).
• Nodes — Kubernetes Nodes with scheduling, metrics, and VMI counts.
• Topology — Interactive node ↔ VM graph (vis-network).
• Export — CSV / XLSX downloads of VM inventory.
• Metrics — App exposes GET /metrics in Prometheus text format for scraping.

Based on / inspired by

Ideas and goals here overlap with tools that help operators see, export, and reason about virtual infrastructure:

• RVTools — the well-known VMware vSphere inventory and reporting workflow (spreadsheet-friendly exports, environment-wide visibility).
• OVTools (OpenShift Virtualization Tools) — RVTools-style consolidated inventory and reporting for OpenShift Virtualization / Kubernetes-native VMs across namespaces, rather than stitching together kubectl / YAML by hand.
• KubeVirt Manager — a web UI focused on operating KubeVirt day to day (VMs, storage, monitoring, and related resources).

KubeVirt Tools is not affiliated with those projects; it is an independent Phoenix app aimed at similar cluster-wide clarity on plain KubeVirt clusters.

Requirements

• Elixir & Erlang installed.
• mix command available in $PATH
• A cluster reachable with the uploaded kubeconfig (KubeVirt CRDs where applicable)

Quick start

Run locally (development)

git clone https://github.com/z7ealth/kubevirt_tools.git
cd kubevirt_tools
mix setup
mix phx.server

Open http://localhost:4000 and upload a valid kubeconfig.

Deploy on Kubernetes / OpenShift

1. Clone the repository and enter the directory:

   git clone https://github.com/z7ealth/kubevirt_tools
   cd kubevirt_tools

2. Create the namespace (skip if it already exists):

   kubectl create namespace kubevirt-tools

   On OpenShift you can use oc new-project kubevirt-tools instead.

3. Configure secrets — edit deploy/k8s/secret.yaml and set SECRET_KEY_BASE (e.g. mix phx.gen.secret), PHX_HOST, and any optional values such as PROMETHEUS_URL.

4. Apply the manifests (from the repo root):

   kubectl apply -f deploy/k8s/

   With the OpenShift CLI:

   oc apply -f deploy/k8s/

   deploy/k8s/route.yaml is OpenShift-only (route.openshift.io). On a plain Kubernetes cluster, skip that file and expose the app with an Ingress (or another controller) targeting the kubevirt-tools-service Service yourself.

Quick expose (port-forward)

Without a Route or Ingress, you can reach the app by forwarding the Service port (Service port 80 maps to the container’s 4000):

kubectl port-forward -n kubevirt-tools svc/kubevirt-tools-service 4000:80

With the OpenShift CLI: oc port-forward -n kubevirt-tools svc/kubevirt-tools-service 4000:80.

Then open http://localhost:4000. Leave the command running while you use the UI.

Testing

This app has been tested on a KubeVirt cluster and works as expected in that environment.

It may also work on OpenShift Virtualization, but this has not been tested yet.

Work in progress (WIP)

Kubernetes deployment — Sample manifests live under deploy/k8s/ (Deployment, Service, RBAC, Secret template). A generic Ingress manifest is not included; use the Route only on OpenShift (see Deploy on Kubernetes / OpenShift above).

Build

To produce a production OTP release (self-contained under _build/prod/rel/kubevirt_tools/), from the repository root:

MIX_ENV=prod mix deps.get --only prod
MIX_ENV=prod mix assets.deploy
MIX_ENV=prod mix release --overwrite

Omit --overwrite on the first build; use it when rebuilding so mix release does not stop at an interactive prompt.

Set runtime environment variables as in Configuration (notably SECRET_KEY_BASE in :prod). Start the release with the HTTP server enabled:

cd _build/prod/rel/kubevirt_tools
export SECRET_KEY_BASE="…"   # e.g. output of mix phx.gen.secret
export PHX_HOST=example.com  # public hostname for URL generation
export PORT=4000
PHX_SERVER=true bin/kubevirt_tools start

The release must be built on an OS compatible with the machine where you run it (same distribution family / libc expectations as a typical Elixir release). For more context, see the Phoenix deployment guide.

---

Configuration

Environment variables

Variable	Required	Default	Purpose
PORT	No	4000	HTTP listen port (see config/runtime.exs).
PHX_SERVER	For releases	—	If set (any value), enables the web server (server: true on the endpoint). Example: PHX_SERVER=true bin/kubevirt_tools start.
SECRET_KEY_BASE	Yes in :prod	—	Secret for signing cookies and tokens. Generate with mix phx.gen.secret.
PHX_HOST	No (:prod)	example.com	Public host used in url: for the endpoint (production).
PROMETHEUS_URL	No	http://localhost:9090	Base URL for the Prometheus HTTP API (/api/v1/query, /-/healthy). Trimmed whitespace; set when Prometheus is not on localhost.
KUBERNETES_INSECURE_SKIP_TLS_VERIFY	No	Unset (verify TLS)	If set to a truthy value (true, 1, yes, on, case-insensitive), the app’s Kubernetes clients skip TLS certificate verification for in-cluster service-account connections and merge with kubeconfig mode. When unset, kubeconfig uploads still follow each context’s insecure-skip-tls-verify flag. Use only when appropriate (e.g. lab clusters with self-signed CAs).

Application config (config :kubevirt_tools, …)

The following are not environment variables today; they live in config/config.exs (and can be overridden in config/runtime.exs, config/dev.exs, or config/prod.exs for your deployment). Values below are the defaults from config/config.exs.

Key	Default	Purpose
:kubeconfig_max_bytes	512_000	Maximum kubeconfig upload size (bytes).
:kubeconfig_connect_timeout_ms	12_000	Timeout for the Kubernetes API reachability check at sign-in.
:prometheus_client_timeout_ms	5_000	Timeout for each Prometheus HTTP client call.
:prometheus_poll_interval_ms	300_000	Interval between full Prometheus snapshots (PromQL + node metrics) pushed to the dashboard.
:prometheus_health_interval_ms	60_000	After a successful snapshot, how often to call Prometheus /-/healthy so the UI drops “Connected” quickly if the server stops responding.

---

Learn more

• Phoenix Framework
• Phoenix guides
• KubeVirt

Known issue: Logo

The logo may appear low-quality or "questionable."

This is a known feature.

No fix is planned.