Not wired
Everything below is sample data. This build has no API behind it — the panel deployment reads real nodes and servers from Pterodactyl. Player counts are the one thing neither build can show: Pterodactyl knows a container is running, not whether anyone is on it, which needs a Server List Ping probe.
Node capacity
Three numbers that are routinely different: what’s committed to servers, what’s actually used, and what physically exists. The gap between them is where outages come from.
Critical
{{ warn.text }}
{{ warn.pct }}
{{ node.memTitle }}
{{ node.sub }}Committed — sum of server limits (overcommit is the hatched zone)
{{ node.committedLabel }}{{ node.scaleLabel }}
Actually used — right now
Committed beyond what exists — an allocation problem: lower limits or move a server. Fix in Pterodactyl.
Using more than expected — a behaviour problem: a server near 100% of its own limit will OOM inside its container before the node notices.
Server
State
Limit
Used
Of its limit
Of physical
{{ r.name }}
{{ r.state }}
{{ r.limit }}
{{ r.used }}
{{ r.pct }}
{{ r.phys }}
{{ node.diskTitle }}
{{ node.diskSub }}{{ node.diskSummary }}
node02 · planned
HP EliteDesk 705 G4Moving modded (6,144 MB) here drops node01 to 63% committed — the single move that ends the overallocation.