Managing VMs and containers

The Guests page is where you start managing every VM and container across all your nodes. There is no host to select first: the list reads the latest inventory, so it opens instantly even when a node is slow.

The guests list

Columns: Name (with the OS logo and Proxmox tags), VMID, Status, Node (links to the node page), IP, Type and Resources. Each row has Console, Start, Reboot and Stop buttons, enabled according to the guest’s status. The list refreshes every few seconds.

• Search matches name, VMID, IP and tags.
• Filters: node, type (qemu / lxc), status and group.
• Bulk actions: tick rows to get a bar with Start, Reboot, Shutdown, Stop, Snapshot, Delete, Assign to group, Mute alerts and Unmute alerts. The live refresh pauses while rows are selected.

Groups

Groups are a display-only way to organize the list; they change nothing in Proxmox.

• Manage groups creates, renames and deletes groups, with a colour, an icon and a sort order.
• Group view shows the list grouped by group, with an Ungrouped section.
• Put guests into a group with the Assign to group bulk action or the Group field on the guest’s General section.

Creating a guest

New guest opens a full-page wizard with six steps. Your progress is kept in the browser, so a reload resumes where you left off (passwords are never stored).

1. Placement: the name/hostname, the Node and the guest type: Container (LXC) or Virtual machine (QEMU/KVM). Pick a node, or Automatic (recommended) (see below).
2. Resources: the VMID, a Configuration (a plan, or Custom to set CPU cores, Memory (MB) and Disk size (GB) yourself), the Customer snapshot limit, the Storage for the root disk and optional Additional disks (storage, size and, for containers, a mount path). A plan fills and locks the cores, memory and disk fields.
3. Operating system:
   • VMs: pick an OS group, then a template version from your QEMU templates. Advanced hardware sets BIOS/firmware (SeaBIOS or UEFI), machine type (i440fx or q35), CPU type and display.
   • Containers: pick a Container template from your LXC templates, set the Root password, Start after creation and the container options (see below).
4. Network: the Network bridge (a physical bridge or an SDN project network) and the IP address: DHCP (automatic) or From zone (static), which allocates the next free address from an IP zone. You can request several IPv4 addresses, one IPv6 address, and Expose ports (inbound NAT) for guests on a private network.
5. Access: for VMs the Root password (the Administrator password for Windows templates) and the Cloud-init user; for every guest the Owner, SSH keys and Firewall presets. Choosing an owner loads their stored SSH keys as pre-checked options; you can paste more keys.
6. Overview: a read-only summary. Create guest submits it; Back still changes anything.

Tip

The root password is optional when at least one SSH key is included. Windows has no SSH key login, so a Windows template always needs an Administrator password of at least 8 characters with upper and lower case letters and a digit.

Exposed ports are added to the node’s NAT configuration but not applied: open the node’s NAT page and apply it to activate them. They require a static (zone) address.

Container options

For containers the Operating system step offers:

• Unprivileged container (on by default, recommended).
• Nesting: run Docker, systemd or nested containers inside.
• keyctl: allow the keyctl() syscall.
• FUSE.
• Extra mounts: allow nfs and/or cifs mounts inside the container.

Tip

Running Docker inside a container? Use an unprivileged container with Nesting and keyctl enabled. The same options exist on the REST API create endpoint (unprivileged, nesting, keyctl, fuse, mount).

Automatic placement

With Automatic (recommended) NexoVirt picks the node for you: it compares free RAM, CPU load and free storage on every node and skips nodes that are offline, in maintenance, without a healthy agent or without room for the guest. The choice is shown in step 1 and checked again when you leave the Resources step. If no node has room, you are asked to pick one manually. Weights and the default are set under Settings > Placement.

Jobs and the live provisioning page

Create, reinstall, delete and power actions run as background jobs: the panel queues the job, shows the expected status right away (for example provisioning, stopped or deleting) and a worker performs the Proxmox calls. A new guest keeps the provisioning status until the inventory confirms it is running, so it never flips to unknown on the way.

After Create guest (and after a reinstall) you land on the live provisioning page:

• A step tracker (for a create: Prepare, Create, Start) with an elapsed-time counter.
• A live log of what the worker is doing, with Follow to keep the newest line in view.
• When the job is done the page opens the guest automatically (after a delete: the guests list).
• When it fails, the page shows the reason with a Retry button.

Clicking a guest that is still provisioning or deleting in the list opens this page as well. The guest page shows a Provisioning… banner with View live progress while a job runs, and a “Last … failed” banner with Retry and Dismiss after a failure.

Note

Jobs need the provisioning worker (php spark provision:work), which the installer runs from cron every minute. If a VM’s OS image is not on the node yet, the create waits for the image download instead of failing.

The guest page

Click a guest’s name to open its page. The header carries Start, Shutdown, Reboot, Stop and Console, plus Migrate for moving the guest to another host. A QEMU agent not installed badge appears on VMs without the guest agent (disk usage and in-guest IP detection then are not available). Below the header, cards show the guest’s CPU, memory, storage and uptime.

A side menu switches between sections:

Section	What it contains
Overview	Details (type/OS, node, VMID, IPs, failover protection, backup schedule) and live CPU, memory, traffic and storage charts. Storage shows each drive with used and total size.
Network	The IP addresses card (Add IP from a zone as a new NIC; Change or Release an address), the Anti-spoofing IP filter and a topology view of the guest’s NICs and addresses.
Disks	The guest’s disks with Add disk and Resize disk (grow only).
Snapshots	Create (optionally with RAM state for running VMs), roll back and delete snapshots.
Backups	Back up now, restore and delete backups. See Backups.
Firewall	Enable the guest firewall, set inbound/outbound default policies, add rules (each owned by Admin or Customer) and apply firewall presets.
Activity	Every lifecycle event of the guest, from NexoVirt and from Proxmox itself, with the real error behind failures. Show full history includes earlier guests with the same VMID.
General	Owner, Group and the Alerts for this guest switch.
Plan	Only for guests on a plan: expiry, traffic and lifecycle actions.
Limits	Apply a plan, or set custom CPU cores, memory and port speed. CPU and RAM changes apply on the next reboot.
OS & reinstall	Reinstall the guest (see below).
Delete	Reset root password and Delete guest.

Changing IP addresses

On the Network section, Change moves an address to another free address of its zone and Release gives it back to the zone. Add IP allocates an extra address from a zone (the next free one, or a specific address) and adds it on a new interface (eth1, eth2, …) with Add as new NIC. Reboot the guest to apply IP changes.

Muting alerts

Turn off Alerts for this guest (General section), or use the Mute alerts bulk action, to exclude a guest from all alerts. Node and host alerts are not affected.

Reset password and delete

• Reset password sets a new root password inside the guest. For VMs this uses the QEMU guest agent, so the VM must be running.
• Delete guest removes the guest and its disks from Proxmox. Zone-allocated IP addresses are released automatically.

Note

With the Enterprise approval workflow turned on (Settings > Approvals, “Require approval for: guest delete”), a delete is not executed right away: it is queued under Approvals, where a second admin approves or rejects it. Nobody can approve their own request.

Reinstall

OS & reinstall wipes the guest and rebuilds it with the same VMID, resources and IP addresses:

• Reinstall (same OS) (VMs): uses the OS the guest was last built with. You confirm by typing the VMID and set a new root password.
• Or pick a different OS: choose an OS group and version, the cloud-init user, owner and SSH keys, set a New root password and press Wipe & reinstall.

Additional disks are not re-created by a reinstall; only the root disk is rebuilt.

Consoles

Console opens an in-browser console:

• Containers, and VMs with a serial console, get a text terminal (xterm.js) with selection, copy and paste and resizing.
• VMs without a serial console (for example Windows) use a graphical noVNC screen. Enable serial console adds a serial port to such a VM; it takes effect after the next reboot. Linux VMs created by NexoVirt always get a serial console.

The connection is proxied through NexoVirt, so the Proxmox session cookie never reaches the browser. The console runs over NexoVirt’s own console proxy, which the installer sets up.

Next steps

• Images & templates for the OS templates behind the wizard.
• Plans & quotas for resource bundles and enforcement.
• Networking & NAT for IP zones and exposing private guests.
• REST API to script guest creation and power actions.