Update documentation to clarify GPU control commands

- Revised README, architecture, CLI, and other documentation to specify that both `Ctrl+C` and `Ctrl+D` now stop the GPU while preserving disk data, correcting previous inaccuracies. - Enhanced access card and setup instructions to reflect the updated command behavior for better user understanding. - Updated tests to ensure the new command behaviors are validated and documented correctly, improving overall clarity in GPU management.
2026-08-21 20:12:03 +03:00
parent a882964ce0
commit f8f8dcc93e
14 changed files with 81 additions and 79 deletions
@@ -125,8 +125,7 @@ def render_access_panel(
    cmds.add_row("hold killer", "gpu-rent hold")
    cmds.add_row("стоп GPU", "gpu-rent stop")
    if tunneled:
-        cmds.add_row("Ctrl+C", "туннель off, GPU жив")
-        cmds.add_row("Ctrl+D", "stop GPU, диски остаются")
+        cmds.add_row("Ctrl+C / Ctrl+D", "stop GPU, диски остаются")
    if not tunneled:
        cmds.add_row("туннель", "gpu-rent tunnel --open")

@@ -220,7 +219,7 @@ def print_access_card(
        for line in mcp_snippet_lines(cfg):
            log(line)
        log("hold: gpu-rent hold   |  stop: gpu-rent stop")
-        log("Ctrl+C — туннель off, GPU жив  |  Ctrl+D — stop GPU")
+        log("Ctrl+C / Ctrl+D — stop GPU (диски остаются)")
        log("")
        return
    from gpu_rent.term import console as default_console
@@ -49,7 +49,7 @@ def _die(exc: BaseException) -> None:
    err(str(exc))
    hint = (
        "Дальше: gpu-rent status · gpu-rent logs · gpu-rent stop "
-        "(Ctrl+C на туннеле GPU не гасит; Ctrl+D гасит)"
+        "(Ctrl+C / Ctrl+D на туннеле гасят GPU, диски остаются)"
    )
    msg = str(exc)
    if "gpu-rent status" not in msg and "Дальше:" not in msg:
@@ -809,7 +809,7 @@ def logs(
 def tunnel(
    open_browser: bool = typer.Option(False, "--open", help="Открыть браузер на 17801"),
 ) -> None:
-    """SSH localhost:17801 -> VM :7801. Ctrl+C — туннель off; Ctrl+D — stop GPU."""
+    """SSH localhost:17801 -> VM :7801. Ctrl+C / Ctrl+D — stop GPU (диски остаются)."""
    try:
        cfg = load_config(require_auth=True)
        state = load_state()
@@ -1052,7 +1052,7 @@ def resize_data(gb: int = typer.Option(..., "--gb", help="Новый разме

 watchdog_app = typer.Typer(
    help=(
-        "Локальный сервис: если туннель умер без Ctrl+C / Ctrl+D / stop — "
+        "Локальный сервис: если процесс туннеля умер без stop (kill/reboot) — "
        "через grace удалить compute. Не путать с idle-killer на VM."
    ),
    no_args_is_help=True,
@@ -423,7 +423,7 @@ def dry_run_plan(checks: list[Check]) -> list[str]:
        "₽: в API нет — смотри панель; диск 24/7 даже после stop",
        f"туннель: localhost:{cfg.swarmui_local_port} -> VM :7801",
        "gpu-rent up --yes создаст GPU + SwarmUI и откроет туннель :17801 "
-        "(Ctrl+C не гасит GPU, Ctrl+D гасит)",
+        "(Ctrl+C / Ctrl+D гасят GPU)",
        "только облако без туннеля: gpu-rent up --yes --no-tunnel",
    ]
    flavor = next((c.detail for c in checks if c.name == "flavor" and c.ok), None)
@@ -2,7 +2,7 @@

 VM idle-killer remains the default safety net. This module is opt-in via
 `gpu-rent watchdog install`: while a tunnel is armed, a scheduled tick stops
-compute if the laptop/process died without Ctrl+C detach or `gpu-rent stop`.
+compute if the laptop/process died without Ctrl+C/Ctrl+D ``stop`` or `gpu-rent stop`.
 """

 from __future__ import annotations
@@ -189,7 +189,7 @@ def touch_heartbeat() -> None:


 def detach_lease_keep_gpu() -> None:
-    """Ctrl+C on tunnel: leave GPU running; local tick must not stop."""
+    """Mark the lease detached so the local tick will not stop compute."""
    lease = load_lease()
    if lease is None:
        lease = LocalLease()
@@ -271,7 +271,7 @@ def install_watchdog(
    log(
        f"local-watchdog установлен ({platform}): тик каждые {interval} мин. "
        f"Grace {grace_seconds() // 60} мин после смерти процесса туннеля → stop. "
-        f"Ctrl+C на туннеле GPU не гасит; Ctrl+D — stop. "
+        f"Ctrl+C / Ctrl+D на туннеле — stop. "
        "При SELECTEL_API_TOKEN — toast каждые BALANCE_NOTIFY_STEP_RUB ₽ (дефолт 200)."
    )
    return marker
@@ -99,7 +99,7 @@ def run_setup(

    do_wd = install_watchdog
    if do_wd is None and confirm:
-        do_wd = confirm("Установить local-watchdog (аварийный stop без Ctrl+C)?")
+        do_wd = confirm("Установить local-watchdog (аварийный stop, если процесс умер)?")
    if do_wd:
        from gpu_rent.local_watchdog import install_watchdog as _install

@@ -1,6 +1,6 @@
 """SSH local forward with Nova watchdog.

-Ctrl+C closes the tunnel and leaves the GPU. Ctrl+D (EOF) runs ``stop``.
+Ctrl+C and Ctrl+D (EOF) both run ``stop`` (disks kept).
 """

 from __future__ import annotations
@@ -242,7 +242,7 @@ def run_tunnel(
    current_host = host
    for loc, rem in forwards:
        log(f"туннель 127.0.0.1:{loc} -> {current_host}:{rem}")
-    log("Ctrl+C — туннель off, GPU жив.  Ctrl+D — stop GPU (диски остаются).")
+    log("Ctrl+C / Ctrl+D — stop GPU (диски остаются).")
    log("watchdog: EXPIRED → unshelve + reconnect")

    server = _start_forwarder(cfg, current_host, forwards)
@@ -255,46 +255,61 @@ def run_tunnel(
    else:
        open_url = f"http://127.0.0.1:{cfg.swarmui_local_port}"

-    from gpu_rent.ready import verify_stack_local
-
-    try:
-        local_checks = verify_stack_local(cfg, log, timeout=90.0)
-        state = load_state()
-        state.notes = dict(state.notes or {})
-        state.notes["stack_local"] = [
-            {"name": c.name, "ok": c.ok, "detail": c.detail} for c in local_checks
-        ]
-        save_state(state)
-    except CloudError as exc:
-        log(f"проверка туннеля: {exc}")
-        raise
-
-    from gpu_rent.access_card import print_access_card
-
-    print_access_card(cfg, tunneled=True, host=current_host)
-
-    if open_browser:
-        webbrowser.open(open_url)
-
-    state = load_state()
-    state.phase = "ready_tunneled"
-    save_state(state)
-
    from gpu_rent.local_watchdog import (
-        detach_lease_keep_gpu,
+        clear_lease,
        start_heartbeat_thread,
        stop_heartbeat_thread,
        watchdog_installed,
    )

-    if watchdog_installed():
-        start_heartbeat_thread()
-        log(
-            "local-watchdog: heartbeat активен — аварийное закрытие "
-            "(не Ctrl+C / не Ctrl+D) → stop после grace"
-        )
+    def _halt_gpu(reason: str) -> None:
+        nonlocal server
+        log(f"{reason} — гашу GPU (диски остаются)")
+        _stop_forwarder(server)
+        server = None
+        stop_heartbeat_thread()
+        clear_lease()
+        if stop_gpu is not None:
+            stop_gpu()
+        else:
+            from gpu_rent.session import cmd_stop
+
+            cmd_stop(cfg, log=log)
+        log("туннель закрыт. GPU остановлен.")

    try:
+        if watchdog_installed():
+            start_heartbeat_thread()
+            log(
+                "local-watchdog: heartbeat активен — kill/reboot без stop "
+                "→ stop после grace"
+            )
+
+        from gpu_rent.ready import verify_stack_local
+
+        try:
+            local_checks = verify_stack_local(cfg, log, timeout=90.0)
+            state = load_state()
+            state.notes = dict(state.notes or {})
+            state.notes["stack_local"] = [
+                {"name": c.name, "ok": c.ok, "detail": c.detail} for c in local_checks
+            ]
+            save_state(state)
+        except CloudError as exc:
+            log(f"проверка туннеля: {exc}")
+            raise
+
+        from gpu_rent.access_card import print_access_card
+
+        print_access_card(cfg, tunneled=True, host=current_host)
+
+        if open_browser:
+            webbrowser.open(open_url)
+
+        state = load_state()
+        state.phase = "ready_tunneled"
+        save_state(state)
+
        if wait is not None:
            wait()
            return
@@ -303,17 +318,7 @@ def run_tunnel(
        next_poll = time.time() + poll_seconds
        while True:
            if end_poll(1.0):
-                log("Ctrl+D — гашу GPU (диски остаются)")
-                _stop_forwarder(server)
-                server = None
-                stop_heartbeat_thread()
-                if stop_gpu is not None:
-                    stop_gpu()
-                else:
-                    from gpu_rent.session import cmd_stop
-
-                    cmd_stop(cfg, log=log)
-                log("туннель закрыт. GPU остановлен.")
+                _halt_gpu("Ctrl+D")
                return
            if not server.is_active:
                next_poll = 0
@@ -350,8 +355,7 @@ def run_tunnel(
                    log(f"unshelve/reconnect fail: {exc}")
                    return
    except KeyboardInterrupt:
-        detach_lease_keep_gpu()
-        log("Ctrl+C — туннель закрыт. GPU жив.")
+        _halt_gpu("Ctrl+C")
    finally:
        stop_heartbeat_thread()
        _stop_forwarder(server)
@@ -49,9 +49,9 @@ def cost_and_risk_lines(cfg: Config, *, spot: bool, flavor_name: str) -> list[st
        f"{cfg.idle_minutes} мин пустой очереди → delete compute. Отложить: gpu-rent hold",
        "preemptible: хостер может усыпить (~24 ч окно) → EXPIRED; tunnel сам unshelve, "
        "или gpu-rent up",
-        "Ctrl+C на tunnel — туннель off, GPU жив. Ctrl+D — stop GPU (диски остаются). "
-        "Опционально: gpu-rent watchdog install — аварийное закрытие окна/ребут "
-        "после grace тоже stop (Ctrl+C по-прежнему detach)",
+        "Ctrl+C / Ctrl+D на tunnel — stop GPU (диски остаются). "
+        "Опционально: gpu-rent watchdog install — kill/reboot окна "
+        "после grace тоже stop",
    ]