FAQ

FloppyOps runs on a separate Linux management server with PHP 8.4 (Debian 13+ native, or Debian 12 via the auto-enabled sury.org repo). Minimum 4 GB RAM (8 GB recommended) and 500 MB free disk space. Managed Proxmox servers: PVE 8.x or 9.x, plus PBS optionally — Python 3.9+ on the nodes for the FloppyOps agent.

For basic features (dashboard, monitoring, VM control via PVE API) — no, the PVE API is enough. For updates, snapshots, terminal, ZFS maintenance and PBS-related operations a lightweight Python agent is installed automatically on first connect to each node (one click, no manual setup).

Yes. FloppyOps organises servers by Location → Cluster → Node. As long as the management server can reach the Proxmox API (port 8006), the physical location does not matter. Pro and higher tiers have unlimited locations.

Community is free forever and includes the core management features for up to 3 PVE servers, 1 PBS, 1 location and 3 users — including automatic update/ZFS schedules. Pro raises the limits to 10/3/unlimited/unlimited and adds ZFS replication, custom alerts, Microsoft 365 mail and audit CSV export. Enterprise (75/15/unlimited) adds PXE bare-metal deployment, custom branding, advanced audit exports and support logging. MSP (250/75/unlimited) adds white-label re-branding plus a dedicated contact; extended support options (SLA) are in preparation. How support works per tier is on the support page. Full feature list: see the pricing section on the home page.

• Your data stays — nothing gets deleted
• 14-day grace period: all features stay active so you can renew without disruption
• After the grace period: fall back to the free Community tier (3 servers, 1 PBS, 3 users)
• Pro / Enterprise / MSP features get disabled, no further updates
• Renew anytime — all features come back instantly

The Community version is free forever and includes all core management features (dashboard, VM/CT control, updates, snapshots, backup view, ZFS overview, UPS, QDevice, custom SSL). Test the complete workflow before deciding on a paid tier. Upgrading only unlocks higher limits and additional features like custom alerts, replication, PXE deployment, branding and the advanced audit exports.

Updates can be installed directly from the UI (Settings > App > "Install now") or via CLI. FloppyOps checks the release server (configurable channel: auto / stable / beta+RC, and check interval), downloads the update, creates an automatic backup, runs the migration and verifies the result. Rollback to a previous backup is possible at any time. Pro+ can schedule auto-updates per day and time.

Yes. On the app server (via SSH, in the install directory) the CLI uses the same engine as the UI button and creates a config backup before every update:

• Latest version: sudo bash build/update.sh
• Specific version: sudo bash build/update.sh --version X.Y.Z
• Offline / air-gapped (from a local ZIP, verifies the signature): sudo bash build/update.sh --zip /pfad/floppyops-X.Y.Z.zip --version X.Y.Z
• Roll back to the latest backup: sudo bash build/rollback.sh

Yes. Cluster create + join can be triggered from inside the app (pvecm via SSH, key exchange handled automatically). Nodes are grouped by cluster; backup schedules, storage overview, VM migration and HA resources work cluster-wide. Standalone servers can be managed in parallel to clusters.

FloppyOps has a built-in disaster-recovery flow. Config-Backup ZIPs (database + settings + branding + certificates) can be scheduled (daily/weekly/monthly + HH:MM) and downloaded from the UI. To restore on a fresh server: sudo bash setup.sh --restore /path/to/backup.zip — the setup script installs the dependencies and restores everything including the encryption key. Works across hosts (the DB host can be different).

FloppyOps has self-healing for failed updates: it automatically frees emergency disk space, runs dpkg --configure -a + apt -f install, retries the upgrade and always runs cleanup (autoremove/autoclean). Active for PVE nodes, PBS servers and replication targets. Whatever cannot be repaired surfaces as an issue in the header drawer and the mail digest.

Beta-testers see new releases first via the internal "tester" channel and provide approve/issue feedback directly from the app, before the release is promoted to public beta. The program is invitation-based and requires an NDA. Contact us via the feature-request page or support if you want to participate.

Server credentials are encrypted at rest (AES-256-CBC). All connections use TLS. Role-based access control (admin/operator/viewer) plus rate-limited login. Code integrity is verified at runtime against an RSA-signed manifest; the license-validate response is Ed25519-signed to block MITM downgrades. The Python agent runs with minimal permissions and authenticates via unique per-node tokens.

No. FloppyOps connects outbound from the management server to your Proxmox nodes — PVE API on 8006, PBS on 8007, the FloppyOps agent on 7700 and SSH on 22 (SSH only for initial setup). Everything works over your LAN or VPN; your servers do not need to be reachable from the internet, and the management UI itself does not need to be public either. An agentless relay option (outbound-only, no open inbound port on either side) is on the roadmap.

FloppyOps is fully self-hosted and runs entirely in your own infrastructure. The only outbound traffic to floppyops.com is license validation, the update check and an integrity status report — no server data, metrics, VM details or credentials ever leave your network. No telemetry, no cloud account.