Implement UP_STOP_ON_FAIL option to manage GPU state on installation failure

- Added a new configuration option `UP_STOP_ON_FAIL` to control whether the GPU should be stopped automatically if the `up` command fails, enhancing user control over resource management.
- Updated the CLI to include a `--keep-on-fail` flag, allowing users to prevent GPU shutdown during installation errors.
- Enhanced the installation scripts and documentation to reflect these changes, providing clearer guidance on the new behavior and configuration options.
- Improved error handling in the CLI to ensure proper cleanup of resources in case of failure, preventing unexpected billing for unused GPU resources.

src/gpu_rent/cli.py

@@ -57,6 +57,21 @@ def _die(exc: BaseException) -> None:
     raise typer.Exit(1)
 
 
+def _stop_after_failed_up(cfg, cause: BaseException) -> None:
+    """Delete compute after a failed up so billing does not continue unnoticed."""
+    warn(
+        "up упал — гашу GPU (UP_STOP_ON_FAIL; оставить: --keep-on-fail / UP_STOP_ON_FAIL=false)"
+    )
+    try:
+        cmd_stop(cfg, no_pull=True, log=log)
+        ok("compute остановлен, диски на месте")
+    except Exception as stop_exc:
+        err(
+            f"auto-stop не удался: {stop_exc} — срочно: gpu-rent stop "
+            f"(причина up: {cause})"
+        )
+
+
 def _print_checks(checks, *, quiet: bool = False) -> int:
     failed = blocking_failed(checks)
     if quiet:
@@ -439,6 +454,11 @@ def up(
         "--no-update",
         help="Не делать git pull SwarmUI и установленных extensions",
     ),
+    keep_on_fail: bool = typer.Option(
+        False,
+        "--keep-on-fail",
+        help="Не гасить GPU если up упал (по умолчанию stop; UP_STOP_ON_FAIL=false)",
+    ),
     verbose: bool = typer.Option(
         False,
         "--verbose",
@@ -458,6 +478,8 @@ def up(
     ),
 ) -> None:
     """Create/unshelve GPU; SwarmUI и/или LLM; по умолчанию туннель."""
+    cfg = None
+    up_ok = False
     try:
         from dataclasses import replace
 
@@ -680,6 +702,7 @@ def up(
             ask=None if yes else ask,
             log=log,
         )
+        up_ok = True
         if no_tunnel:
             from gpu_rent.access_card import print_access_card
 
@@ -705,7 +728,24 @@ def up(
             open_browser=open_browser,
             log=log,
         )
+    except KeyboardInterrupt as exc:
+        if (
+            not up_ok
+            and cfg is not None
+            and not keep_on_fail
+            and bool(getattr(cfg, "up_stop_on_fail", True))
+        ):
+            _stop_after_failed_up(cfg, exc)
+        err("прервано (Ctrl+C)")
+        raise typer.Exit(130) from exc
     except GpuRentError as exc:
+        if (
+            not up_ok
+            and cfg is not None
+            and not keep_on_fail
+            and bool(getattr(cfg, "up_stop_on_fail", True))
+        ):
+            _stop_after_failed_up(cfg, exc)
         _die(exc)