See hardware health and bandwidth at a glance

The FlexDedi panel watches the layers you can't see well from inside the OS: the hardware and the network port. Bandwidth graphs show inbound and outbound traffic per server, hardware health shows whether components are reporting problems, and email alerts make sure an issue doesn't wait for you to log in.

Everything above the hardware is yours to monitor, because servers are unmanaged. A practical baseline:

• Metrics: Prometheus with node_exporter, or your preferred agent, for CPU, memory, disk and network inside the OS.
• Disks: smartctl and nvme smart-log for drive wear and errors, plus /proc/mdstat or zpool status for RAID health.
• Services: health checks on the ports and endpoints your users depend on, run from outside the server.
• Logs: the systemd journal shipped to a central store over the private network.

The two views complement each other: the panel covers hardware and port-level issues, and your own monitoring covers the OS and your applications.

Bandwidth graphs answer practical questions quickly:

• Is the port the bottleneck? Sustained traffic close to the port speed (10, 25, 2 × 25 or 2 × 100 Gbps depending on the server) means it is time to add servers or move to a server with a faster port.
• Is this an attack or a launch? A jump in inbound traffic without a matching rise in requests or outbound traffic deserves a look. Always-on DDoS protection filters floods at the network edge, and our engineers can investigate with you if service is affected.
• Is something leaking data? Unexpected outbound traffic at odd hours can mean a compromised service or a misconfigured backup job.

Traffic is unmetered on every server, so the graphs are for visibility, not billing: there is no monthly cap to watch and no overage to avoid. For per-process or per-connection detail, use tools such as iftop or nload, or your own Prometheus metrics.