llmgw/internal/web/locales/en/node.json

{
  "node.last_seen": "last seen %s ago",
  "node.never_seen": "never seen",
  "node.nav.cluster": "cluster",
  "node.cores": "%d cores",
  "node.ram_in_use_of": "RAM in use · of %s",
  "node.load": "load · %s 5m · %s 15m",
  "node.inlet": "inlet air",
  "node.not_measured": "CPU and RAM not measured lately",
  "node.disk_free": "disk free · of %dG",
  "node.gpus_free": "GPUs free",

  "node.gpus": "GPUs",
  "node.gpu.busy": "Busy",
  "node.gpu.memory_used": "Memory used",
  "node.gpu.temp": "Temperature",
  "node.gpu.memory_temp": "Memory temp.",
  "node.gpu.power": "Power",
  "node.gpu.clock": "Clock",
  "node.gpu.held_back": "Held back by, and wear",
  "node.gpu.memory_util": "memory %s%%",
  "node.of_gb": "of %s GB",
  "node.of": "of %s",
  "node.gpu.nothing": "nothing",
  "node.gpu.ecc": "%s uncorrected ECC errors",
  "node.gpu.pcie": "%s PCIe retries",

  "node.fabric": "Fabric",
  "node.fabric.on": "On %s",
  "node.fabric.map": "the map →",
  "node.fabric.none": "On no fabric: no port a declared fabric names can carry traffic, so no multi-node replica is placed here.",
  "node.ib.adapter": "Adapter",
  "node.ib.rate": "Rate",
  "node.ib.physically": "Physically on",
  "node.ib.errors": "Errors since boot",
  "node.ib.sent": "Sent",
  "node.ib.received": "Received",
  "node.unknown": "unknown",
  "node.none": "none",
  "node.ib.note": "Re-read on every heartbeat. A port that goes down takes its fabric off this node within seconds; one that comes back returns it. Errors are cumulative since the adapter came up: a number that grows is the one to look at.",

  "node.nvl.title": "GPUs and NVLink",
  "node.nvl.domain": "In NVLink domain <code>%s</code>: tensor parallelism can span this node and the rest of its domain.",
  "node.nvl.ends_here": "NVLink ends at this machine.",
  "node.imex.active": "active",
  "node.imex.channels.one": "1 channel",
  "node.imex.channels.other": "%d channels",
  "node.imex.not_installed": "not installed",
  "node.nvl.matrix": "NVLink between GPUs",
  "node.nvl.over_nvlink": "GPU %d ↔ GPU %d over NVLink",
  "node.nvl.over_pcie": "GPU %d ↔ GPU %d over PCIe",
  "node.gpu.name": "Name",
  "node.gpu.compute": "Compute",
  "node.gpu.numa": "NUMA",
  "node.gpu.nearest": "Nearest adapter",
  "node.gpu.multinode": "Multi-node NVLink",
  "node.gpu.joined": "joined",
  "node.gpu.bandwidth": "%s bandwidth",
  "node.nvl.note": "NV is a pair of cards joined by NVLink; a tensor-parallel group kept to NV pairs runs its all-reduces at NVLink speed, and placement picks it that way. The nearest adapter is where a card's share of traffic between nodes should go: PIX and PXB are under the same PCIe switch, SYS is across the processors.",

  "node.runs": "What runs here",
  "node.memory": "Memory",
  "node.cores_f": "%s cores",
  "node.no_engines": "No engines are placed on this node.",
  "node.version_differs": "the control plane runs %s",
  "node.gpu.models": "Models",
  "node.engines": "Engines: %s",
  "node.engines_none": "Engines: none installed",
  "node.model_files": "%d model file(s) on disk",

  "node.last": "The last %s",
  "node.chart.cpu_note": "All cores together, from the node's heartbeat.",
  "node.chart.ram": "RAM in use",
  "node.chart.ram_note": "Total less what the kernel could hand to a new process.",
  "node.chart.cores": "Cores busy, by model",
  "node.chart.cores_note": "Each engine's process group, the way top counts it.",
  "node.chart.engine_memory": "Engine memory, by model",
  "node.chart.engine_memory_note": "Resident memory, including the weights mapped from disk. The kernel can take those pages back, so this can exceed RAM in use.",
  "node.chart.gpu_busy": "GPU busy, by card",
  "node.chart.gpu_busy_note": "Of each sample period, the share a kernel was running — not how much of the chip it used: one SM busy reads 100%. A card held by a replica and idle here is one nobody is asking.",
  "node.chart.gpu_power_load": "GPU power load, by card",
  "node.chart.gpu_power_load_note": "Draw against the card's power limit. Busy says a kernel ran; this rises with how much of the chip it ran on.",
  "node.chart.gpu_sm_active": "SMs active, by card",
  "node.chart.gpu_sm_active_note": "The share of time the card's SMs had work, from its profiling counters. When busy reads 90 and this 60, a third of the chip sits idle while a kernel runs.",
  "node.chart.gpu_memory": "GPU memory used, by card",
  "node.chart.gpu_memory_note": "Framebuffer in use: the weights and the KV cache an engine reserved, and anything not llmgw's.",
  "node.chart.gpu_temp": "GPU temperature, °C",
  "node.chart.gpu_temp_note": "Throttling usually starts in the high 80s; see what holds a card back in the table above.",
  "node.chart.gpu_power": "GPU power, W",
  "node.chart.gpu_power_note": "Draw against the card's enforced limit; at the limit is sw_power_cap.",
  "node.chart.load": "Load average",
  "node.chart.load_note": "Above the core count means work is waiting for a CPU.",
  "node.chart.queue_note": "Requests accepted and not yet started, on the replicas here now.",
  "node.no_store": "No metrics store is configured, so there is no history here; <code>monitoring.prometheus</code> in the server configuration adds it.",

  "logs.title": "%s logs",
  "logs.agent": "the agent's log",
  "logs.engine": "output of engine <code>%s</code>",
  "logs.last_lines": "last %d lines",
  "logs.lines": "lines",
  "logs.stop": "stop following",
  "logs.follow": "follow",
  "logs.follow_every": "(refresh every 5s)",
  "logs.empty": "Nothing logged yet.",
  "logs.cannot_read": "this control plane cannot read node logs",

  "nodeengine.notice.auto": "%s installs %s in the build its hardware runs; the node is told now, and installs it in the background",
  "nodeengine.notice.none": "%s installs %s no longer; the node is told now, and installs it in the background",
  "nodeengine.notice.build": "%s installs %s built for %s; the node is told now, and installs it in the background"
}