feat(labsim): live topology view with per-path latency

The Grafana heatmap of 1s and 0s said almost nothing, and the state timeline
was an unreadable pile of overlapping series labels. Replaced as the primary
view with a purpose-built page served by the exporter itself.

- Probe now captures ICMP RTT, exposed as labsim_rtt_ms{src,dst}. A path that
  is up but slow is a different problem from one that is down, and a pass/fail
  grid cannot show it.
- Exporter serves / (topology), /api/matrix (JSON) and /metrics.
- topology.html: node per VLAN in a ring, VyOS router in the centre because
  every inter-VLAN packet really does traverse it, one line per pair coloured
  green/red with the RTT on it. Hovering gives per-direction state. A node ring
  goes red if anything to or from it is blocked. Side panels list blocked paths
  and the slowest links. Refreshes every 5s, no dependencies.

Grafana stays for what it is actually good at — history of when a path flipped.

Label placement is deliberate: RTT captions sit ~32% along each edge with a
perpendicular nudge, because every diagonal of a 6-node mesh crosses the centre
and midpoint labels stack on the router node.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DMVzWZgiKW2wquf5z8S1yH

labsim/labsim-matrix.py

@@ -37,18 +37,25 @@ PROTOS = ("icmp", "tcp22", "tcp80")
 # Runs ON the guest. Keep it stdlib-only and quick — a hung probe delays the
 # whole sweep, so every check is hard-bounded by a timeout.
 PROBE = r'''
-import json, socket, subprocess, sys
+import json, re, socket, subprocess, sys
 targets = json.load(sys.stdin)
 out = {}
 for name, ip in targets.items():
     res = {}
     try:
-        res["icmp"] = subprocess.run(
-            ["ping", "-c", "1", "-W", "1", ip],
-            stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, timeout=4
-        ).returncode == 0
+        p = subprocess.run(["ping", "-c", "1", "-W", "1", ip],
+                           stdout=subprocess.PIPE, stderr=subprocess.DEVNULL, timeout=4)
+        res["icmp"] = p.returncode == 0
+        # RTT as well as pass/fail: a path that is up but slow is a different
+        # problem from one that is down, and the grid alone cannot show it.
+        res["rtt_ms"] = None
+        if res["icmp"]:
+            m = re.search(r"time[=<]\s*([0-9.]+)\s*ms", p.stdout.decode("utf-8", "replace"))
+            if m:
+                res["rtt_ms"] = float(m.group(1))
     except Exception:
         res["icmp"] = False
+        res["rtt_ms"] = None
     for port in (22, 80):
         s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
         s.settimeout(1.5)
@@ -147,7 +154,7 @@ def render(vlans: list[dict], results: dict, prev: dict | None, protos: tuple[st
             print(row)
 
     reach = sum(1 for s in results.values() if "__error__" not in s
-                for d in s.values() for p in protos if d.get(p))
+                for d in s.values() for p in protos if d.get(p) is True)
     total = sum(1 for s in results.values() if "__error__" not in s
                 for _d in s.values() for _p in protos)
     print(f"\n  reachable: {reach}/{total}   "