0.0s  Linux; real T3 servers; fresh independent homes; real Tailscale CLI via sudo
   0.0s  Start A on 24461, HTTPS Serve port 24463
   9.2s  A log: Tailscale Serve configured
   9.3s  A started: Serve :24463 / -> http://127.0.0.1:24461
   9.3s    A loopback=200; B loopback=unreachable; tailnet HTTPS=200
   9.3s  Start B on 24462 with the same Serve port
  16.5s  B log: Tailscale Serve configured
  16.5s  B started: Serve :24463 / -> http://127.0.0.1:24462
  16.5s    A loopback=200; B loopback=200; tailnet HTTPS=200
  16.5s  SIGKILL B (unclean exit)
  16.6s  Immediately after killing B: Serve :24463 / -> http://127.0.0.1:24462
  16.6s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  21.6s  After B exit +5s: Serve :24463 / -> http://127.0.0.1:24462
  21.6s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  26.6s  After B exit +10s: Serve :24463 / -> http://127.0.0.1:24462
  26.6s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  31.6s  After B exit +15s: Serve :24463 / -> http://127.0.0.1:24462
  31.6s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  36.7s  After B exit +20s: Serve :24463 / -> http://127.0.0.1:24462
  36.7s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  41.8s  After B exit +25s: Serve :24463 / -> http://127.0.0.1:24462
  41.8s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  46.9s  After B exit +30s: Serve :24463 / -> http://127.0.0.1:24462
  46.9s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  52.0s  After B exit +35s: Serve :24463 / -> http://127.0.0.1:24462
  52.0s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  57.0s  After B exit +40s: Serve :24463 / -> http://127.0.0.1:24462
  57.0s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  57.0s  Set Serve explicitly to dead 24462 to exercise stale-route recovery on both commits
  57.1s  Stale route injected: Serve :24463 / -> http://127.0.0.1:24462
  57.1s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  62.2s  After stale route +5s: Serve :24463 / -> http://127.0.0.1:24462
  62.2s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  67.3s  After stale route +10s: Serve :24463 / -> http://127.0.0.1:24462
  67.3s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  72.3s  After stale route +15s: Serve :24463 / -> http://127.0.0.1:24462
  72.3s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  77.4s  After stale route +20s: Serve :24463 / -> http://127.0.0.1:24462
  77.4s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  82.5s  After stale route +25s: Serve :24463 / -> http://127.0.0.1:24462
  82.5s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  87.5s  After stale route +30s: Serve :24463 / -> http://127.0.0.1:24462
  87.5s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  92.5s  After stale route +35s: Serve :24463 / -> http://127.0.0.1:24462
  92.5s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  97.6s  After stale route +40s: Serve :24463 / -> http://127.0.0.1:24462
  97.6s    A loopback=200; B loopback=unreachable; tailnet HTTPS=502
  97.6s  End of observation; stop A cleanly and remove only the dedicated Serve route