Files
cv4pve-cli-tui/README.md
T
DuckbootandClaude Opus 5.5 0df928ab11
CI / pytest and ruff (pull_request) Successful in 7m25s
CI / pytest and ruff (push) Successful in 6m56s
Restore a guest from a backup in the info screen
r on the info screen restores the backup under the cursor of the
backup list.

- New guest (default): the next free VMID from `get cluster nextid`,
  editable. The original is left untouched. An id already used by a
  listed guest is refused.
- Overwrite: replaces the guest itself. The guest must be stopped; its
  live status is read at apply and again right before the call. The
  VMID must be typed to confirm, then an "Overwrite" confirm follows.

--storage is always passed. It is preselected to the storage of the
guest's rootfs or first disk, and the choices are the node's storages
that can hold CT (rootdir) or VM (images) disks. Start after restore
is off by default, because a copy keeps the original's MAC address
and IP.

Safety:
- cli.restore refuses force (overwrite) for any VMID other than the
  guest's own (explicit raise). A hypothesis property checks that no
  other VMID ever reaches the CLI with force.
- The volid and storage are validated before they become arguments.
- A backup of the other guest type is refused. A reused VMID can carry
  a qemu backup for a CT or the other way round, and whether PVE
  rejects a mismatched archive before removing the old guest is not
  verified.
- An earlier guest's backup (visible with a) is labelled with that
  guest's name in the dialog and the confirm.
- r before the info screen has loaded says "still loading".
- Restore gets a 1-hour CLI timeout.

Not run live: a live test would create a guest that then has to be
deleted, which the issue reserves for the user's approval. A
read-only check opened the dialog for CT 104 (next id 105, local-lvm
preselected) and closed it without any call.

Closes #15

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 01:32:05 +01:00

4.8 KiB

cv4pve-cli-tui

CI

Textual TUI for cv4pve-cli: list Proxmox VE guests (VMs and LXC), start/shutdown/stop/reboot them, and change cores, memory and disk size. The table shows CPU % and Mem % of running guests and the info screen shows in-use values (refresh with g).

It shells out to cv4pve-cli and uses its current context (cv4pve-cli config use <name>); it stores no credentials itself. Tested with cv4pve-cli 2.4.0; with another version the app warns at startup, because it parses output formats that are not a stable API.

Install

Needs uv and a working cv4pve-cli on PATH.

uv tool install git+https://git.duckboot.net/Duckboot/cv4pve-cli-tui
cv4pve-tui                      # from any directory
uv tool upgrade cv4pve-tui      # later: get the newest main
uv tool uninstall cv4pve-tui

From a local checkout instead: uv tool install . (add --force to reinstall after changes).

Run

Installed: cv4pve-tui. From a checkout without installing:

uv run cv4pve-tui
Key Action
s start
S shutdown (confirm)
x stop - hard power-off (confirm)
r reboot (confirm)
Enter info: resources (CPU, memory, swap, disks), backups, backup job, network (config and runtime IPs)
Enter then e in info: edit cores / memory / grow a disk (disk growth asks again; it cannot be undone)
Enter then b in info: backup now (snapshot, zstd, storage choice); a there = also show backups of earlier guests that had the same VMID
Enter then n in info: edit network (bridge, VLAN, firewall, IPv4/IPv6; VM IP via cloud-init at next boot); u there = undo the last network change of this guest (this session only)
Enter then o in info: edit autostart, startup order (order, up and down delay) and tags (one confirmed set config call; empty fields remove the setting)
Enter then p / R / D in info: take a snapshot (name, description, VM: include RAM) / roll back to the snapshot under the cursor (confirms; starts the guest again if it is running) / delete the snapshot under the cursor (confirms)
Enter then f in info: fstrim in a running VM with the guest agent enabled (confirms; containers: use pct fstrim on the node)
Enter then r in info: restore the backup under the cursor, as a new VMID (default, the original stays untouched; a copy keeps the original's network settings, so leave Start off until they are changed) or overwriting the guest itself (stopped guests only; type the VMID, then confirm)
t shell in a running LXC container (ssh -t root@<node> pct enter <vmid>)
l tasks: recent cluster tasks (start, duration, node, type, VMID, user, status; ! marks failures); Enter shows the log of the task under the cursor, g reloads. In info: l = only this guest's tasks
U package updates: pending updates per online node (node, package, current, new, origin, priority), read-only; g reloads. The list is as of the node's last apt update. The frame's bottom border shows the count
g refresh
/ filter the guest table as you type: space-separated terms, all must match (case-insensitive); plain term = part of name, VMID or a tag; is:running / is:stopped; type:lxc (or ct) / type:vm (or qemu). Enter keeps the filter and returns to the table, Escape clears it. The frame's bottom border shows filter: <text> · N of M
c switch cv4pve-cli context (cluster)
q quit

Warning: link changes (bridge, VLAN, firewall) apply immediately on a running guest and can cut its network (and DNS for others if the guest is your DNS server). IP changes on a running container may need a reboot of the container (seen: static -> DHCP gave no IPv4 until a reboot).

The shell key needs key-based root SSH to the PVE node (LXC only).

Switching context runs cv4pve-cli config use, which changes cv4pve-cli's global current context: other shells and scripts follow, and it stays after the app exits.

A ! in the first column marks a guest that no backup job covers (Proxmox's not-backed-up list); the frame's bottom border shows how many.

Empty fields in the edit dialog are left unchanged. If a CPU/memory change cannot be hot-plugged, the app tells you it applies at the next reboot.

Development

uv sync
uv run pytest
uv run ruff check src tests && uv run ruff format --check src tests

Tests mock cv4pve-cli; they never touch a cluster. CI (.gitea/workflows/ci.yaml) runs the same commands on every PR into main; ruff is pinned in the dev group so local runs and CI check the same rules. Design: docs/plans/guest-tui-mvp.md.