docs: promote /web from develop for backup-restore, Log2RAM and Network Flow guides

Brings 69 files from develop under /web/:
- Full Backup & Restore section (11 pages EN + ES: overview, how-it-works, destinations,
  creating backups, scheduled jobs, restoring, cross-kernel hydration)
- Log2RAM dedicated block in post-install/optional with commands + upstream link
- Network Flow diagram documented on monitor/dashboard/network
- Rewritten category descriptions in post-install/customizable
- Fixed automated.json thresholds + link to Log2RAM section
- Updated screenshots (network-flow-overview, storage-top-row, vms modals)

No code, config or AppImage binaries touched — /web/ scope only. Merging deploys
the documentation site to the current beta release notes.

web/messages/en/docs/monitor/dashboard/network.json

@@ -35,6 +35,32 @@
       }
     ]
   },
+  "flow": {
+    "heading": "Network Flow diagram",
+    "intro": "Between the top row and the three group cards, the tab renders a live topology view called <strong>Network Flow</strong>. It draws every path a packet can take through the host — physical NICs at one end, bridges in the middle, VMs and containers at the other — with animated pulses that show real-time rx/tx traffic on each link.",
+    "imageAlt": "Network Flow diagram — NICs on the left, host and bridges in the middle, VMs and LXCs on the right, with animated pulses on each connection",
+    "imageCaption": "Network Flow — colour-coded nodes plus animated comets whose direction and thickness reflect live traffic.",
+    "elementsTitle": "What the diagram shows",
+    "elementsIntro": "Every node is one entity from the same inventory used by the group cards below, but arranged as a topology so relationships between them are immediate:",
+    "elements": [
+      "<strong>NICs</strong> (amber) — every physical interface with link up. Interfaces reported as down are drawn faded.",
+      "<strong>Host</strong> (amber) — the Proxmox host itself, sitting between the NICs and the bridges. Rendered as a fixed anchor; not clickable.",
+      "<strong>Bridges</strong> (cyan) — Linux and OVS bridges. Only bridges that carry at least one active guest are drawn; unused bridges are hidden to keep the diagram focused on what actually moves traffic.",
+      "<strong>LXCs</strong> (cyan) — running containers attached to a bridge.",
+      "<strong>VMs</strong> (purple) — running virtual machines attached to a bridge.",
+      "Offline guests (stopped VMs / containers) are hidden."
+    ],
+    "pulsesTitle": "What the animation encodes",
+    "pulsesBody": "The animated comets that travel along each edge represent live traffic:",
+    "pulses": [
+      "<strong>Direction</strong> — a pulse flowing toward the NIC is <em>tx</em> from the guest; toward the guest is <em>rx</em>.",
+      "<strong>Stroke width</strong> scales with the guest's combined rx+tx rate. An idle guest still draws a faint animated line; a busy one gets a thick one.",
+      "<strong>Comet head glow</strong> — warm at ~1 MB/s, hot at ≥30 MB/s. Useful to spot at a glance which guest is dominating a NIC.",
+      "<strong>Click</strong> — every node except the host is clickable and opens the same per-interface drill-in modal documented below. Tapping the host is intentionally a no-op — there is no host-level modal in this view."
+    ],
+    "useTitle": "When it's useful",
+    "useBody": "The group cards below tell you <em>what</em> exists; the diagram tells you <em>how it's connected</em> and where the traffic is flowing right now. Patterns easy to spot at a glance: a bridge with no active guest attached, two heavy guests going through the same NIC (bottleneck candidate), or a single VM saturating a link while everything else is quiet."
+  },
   "groups": {
     "heading": "Three interface groups",
     "intro": "Below the top row, three cards split the inventory by role. Each card has its own active-count badge in the header. Interface <strong>type</strong> is identified at a glance by a coloured badge on every row:",

web/messages/en/docs/monitor/dashboard/system-overview.json

@@ -53,6 +53,30 @@
     "sparklineTitle": "The sparkline is meaningful",
     "sparklineBody": "The temperature card draws a 5-minute trace under the value, with the line and gradient colour following the same Warning/Critical pair documented above. It's the fastest way to see whether the host is in a thermal climb without opening the detail modal."
   },
+  "processes": {
+    "heading": "Top processes by CPU / Memory",
+    "intro": "The CPU Usage and Memory cards on the Overview tab are clickable. Clicking either opens a sortable list of the top 25 processes — ordered by CPU usage when opened from the CPU card, by resident memory when opened from the Memory card.",
+    "listTitle": "The list dialog",
+    "listItems": [
+      "<strong>Auto-refresh</strong> — the list updates every 5 seconds while the dialog is open and stops polling as soon as it closes.",
+      "<strong>Filter</strong> — the search box narrows the list by command, user or PID.",
+      "<strong>Inline bar</strong> — the primary metric column draws a small bar scaled to the highest value in the filtered list, so ranking stays visible even when no process is near 100 %.",
+      "<strong>Mobile layout</strong> — under 640 px the PID and User columns hide so Command, CPU % and Memory still fit on a phone screen without horizontal scroll."
+    ],
+    "captureListAlt": "Top processes by Memory modal — table with PID, USER, COMMAND, CPU %, Memory columns sorted by RSS",
+    "captureListCaption": "The Memory card opens the list sorted by RSS (indigo accent). The CPU card opens the same list sorted by CPU usage (blue accent).",
+    "detailTitle": "Per-process detail",
+    "detailIntro": "Clicking any row in the list opens a second dialog with the live picture of that one process, organised in four sections:",
+    "detailItems": [
+      "<strong>Overview</strong> — state, parent process, thread count, open file descriptors, user and group.",
+      "<strong>Resources</strong> — CPU %, Memory %, Resident (RSS), Virtual size, Swap, I/O read and write totals.",
+      "<strong>Command</strong> — short name, full command line, executable path and working directory.",
+      "<strong>Lifetime</strong> — start timestamp and elapsed runtime."
+    ],
+    "detailRefresh": "The detail dialog refreshes every 3 seconds while open. If the process finishes mid-dialog, polling stops, an amber <em>This process has finished</em> banner appears, and the last captured snapshot stays on screen (dimmed) so you can still see what was happening just before it ended.",
+    "captureDetailAlt": "Process detail modal — Overview, Resources, Command and Lifetime sections for a single PID",
+    "captureDetailCaption": "Per-process detail dialog opened from a list row. The accent colour matches the card that opened it (blue for CPU, indigo for Memory)."
+  },
   "middle": {
     "heading": "Middle: node metrics charts",
     "body1": "Below the top row sits the <code>NodeMetricsCharts</code> component — historical CPU, memory and disk-I/O graphs sourced from Proxmox's own RRD store via <code>/api/node/metrics</code>. A timeframe selector switches between <em>1 hour / 24 hours / 7 days / 30 days / 1 year</em>; data resolution drops as the window grows so the chart stays smooth.",