Product
Commercial-grade infrastructure features
One control plane for the work that usually sprawls across panels, tickets, and SSH sessions — mapped to the modules operators actually open.
Fleet & host control
Inventory Linux hosts and remote agents, watch live health, and run day-two operations without leaving the plane.
- Local and remote agent inventory with health and connectivity signals
- Host filesystem, packages, services, and power actions behind fleet grants
- Discovery that maps what is actually running before you change it
Modules: Server Manager · Infrastructure · Monitoring
Websites & applications
Provision sites and application stacks beside the servers that run them — one context from overview to deploy.
- Virtual host and application lifecycle in the same workspace as the host
- Reverse-proxy rules kept next to the sites they protect
- Fewer tool hops between “what is live” and “what to change”
Modules: Websites · Applications · Reverse Proxy
SSL & DNS operations
Issue and renew certificates, keep zones aligned, and reduce expiry surprises across the estate.
- Certificate inventory with renewal-oriented workflows
- DNS zones and records managed as infrastructure metadata
- Signals that connect cert/DNS state to the sites that depend on them
Modules: SSL Center · DNS Center
Data, files & mail
Browse remote filesystems, oversee databases, and treat mail as first-class infrastructure — not a bolted-on afterthought.
- Chunked uploads and host file operations with least-privilege grants
- Database metadata and operations routed through the control plane
- Mail domains, accounts, and DNS suggestions in the same narrative
Modules: File Manager · Database Center · Mail Center · Upload Center
Security & access
Hardening and identity live next to the work they protect — firewall, access, and audit in one plane.
- Firewall, Fail2Ban, geo controls, and scanners from the operator surface
- Passkeys, sessions, RBAC, and per-server fleet grants
- Root vs limited privilege with explicit host access levels
Modules: Security · Users & Access
Automation & AI ops
Backups, workflows, alerts, and an infrastructure-aware copilot that stays close to the controls you use.
- Backup jobs and targets with scheduler-aware settings
- Workflows and webhook triggers for repeatable ops
- AI chat, forecasts, and recommendations grounded in fleet context
Modules: Backup Center · Automation · AI Center · Log Center
What is CanonPanel?
AI-native infrastructure control plane for Linux fleets, websites, SSL, DNS, databases, mail, security, backups, and automation — one composed surface for modern ops.
Is CanonPanel self-hosted or SaaS?
CanonPanel is designed as a self-hosted control plane: Next.js BFF, Express metadata (MySQL), Java OS edge, optional agent, and AI service. You run it on your network; there is no required multi-tenant SaaS dependency for core ops.
Which operating systems are supported?
The panel orchestrates Linux hosts through the Java edge and remote agents. Windows host paths exist for agent/edge scenarios; day-to-day fleet ops target Linux distributions with systemd, packages, and common hosting stacks.
Does CanonPanel replace SSH?
It centralizes day-two operations and host control. SSH remains available when you need a raw shell; the panel is the composed operating surface so most work does not require hopping consoles.
How does access control work?
Express owns identity: Argon2id passwords, JWT sessions in httpOnly cookies, WebAuthn passkeys, RBAC permissions, and per-server fleet grants (read / operate / admin). The Next BFF enforces signed-in access and least privilege before host ops reach Java.
Can MSPs manage multiple client estates?
Fleet inventory, labels, and per-server grants are built for multi-host estates. Agencies and MSPs use one plane with clear ownership boundaries instead of a separate panel per client tool.
How are backups handled?
Host/workload backups run through the Java edge and Backup Center. Separately, Express MySQL metadata should be dumped on a schedule (npm run db:backup) so users, DNS/SSL metadata, and configuration are recoverable.
Where does AI fit?
The AI service powers copilot chat, forecasts, log analysis, and recommendations using fleet context from Express/Java. It does not auto-execute destructive host actions without operator intent.
How do I check availability?
Use the public Status page and /api/status JSON feed for uptime monitors. /api/health remains a lightweight public probe for the Next BFF path to Java health.