Fleet management from a dev machine

[titouanmathis]

Problem

Managing multiple DDEV servers (OVH, Hetzner, client-specific) with no unified way to:

• See all servers and their projects
• Check server health (disk, RAM, uptime)
• Run maintenance across the fleet

Proposed Solution

Add a trafic fleet CLI subcommand that runs from a developer's machine and talks to multiple servers over SSH.

Fleet Config

# ~/.config/trafic/fleet.toml

[[servers]]
name = "ovh-preprod"
host = "preprod.ovh.studiometa.fr"
user = "ddev"
tld = "ovh.previews.studiometa.fr"

[[servers]]
name = "hetzner-01"
host = "hetzner01.studiometa.fr"
user = "ddev"

CLI Surface

trafic fleet status              # List all servers with health metrics
trafic fleet projects            # List all projects across all servers
trafic fleet exec "ddev poweroff" # Run command on all servers
trafic fleet backup              # Trigger backups on all servers
trafic fleet upgrade             # Update trafic agent on all servers

Implementation

• Phase 1: SSH commands (works today, no agent changes)
• Phase 2: Agent exposes /__status__/fleet JSON endpoint for faster queries

---

From exploration-ops-features.md
Priority: P3 — nice-to-have for multi-server ops