nvidia-nat 1.1.0a20251020__py3-none-any.whl

nat/front_ends/fastapi/fastapi_front_end_plugin.py

@@ -0,0 +1,247 @@
+# SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import asyncio
+import logging
+import os
+import sys
+import tempfile
+import typing
+
+from nat.builder.front_end import FrontEndBase
+from nat.front_ends.fastapi.dask_client_mixin import DaskClientMixin
+from nat.front_ends.fastapi.fastapi_front_end_config import FastApiFrontEndConfig
+from nat.front_ends.fastapi.fastapi_front_end_plugin_worker import FastApiFrontEndPluginWorkerBase
+from nat.front_ends.fastapi.main import get_app
+from nat.front_ends.fastapi.utils import get_class_name
+from nat.utils.io.yaml_tools import yaml_dump
+from nat.utils.log_levels import LOG_LEVELS
+
+if (typing.TYPE_CHECKING):
+    from nat.data_models.config import Config
+
+logger = logging.getLogger(__name__)
+
+
+class FastApiFrontEndPlugin(DaskClientMixin, FrontEndBase[FastApiFrontEndConfig]):
+
+    def __init__(self, full_config: "Config"):
+        super().__init__(full_config)
+
+        # This attribute is set if dask is installed, and an external cluster is not used (scheduler_address is None)
+        self._cluster = None
+        self._periodic_cleanup_future = None
+        self._scheduler_address = None
+
+    def get_worker_class(self) -> type[FastApiFrontEndPluginWorkerBase]:
+        from nat.front_ends.fastapi.fastapi_front_end_plugin_worker import FastApiFrontEndPluginWorker
+
+        return FastApiFrontEndPluginWorker
+
+    @typing.final
+    def get_worker_class_name(self) -> str:
+
+        if (self.front_end_config.runner_class):
+            return self.front_end_config.runner_class
+
+        worker_class = self.get_worker_class()
+
+        return get_class_name(worker_class)
+
+    @staticmethod
+    async def _periodic_cleanup(scheduler_address: str,
+                                db_url: str,
+                                sleep_time_sec: int = 300,
+                                log_level: int = logging.INFO):
+        from nat.front_ends.fastapi.job_store import JobStore
+
+        job_store = JobStore(scheduler_address=scheduler_address, db_url=db_url)
+
+        logging.basicConfig(level=log_level)
+        logger.info("Starting periodic cleanup of expired jobs every %d seconds", sleep_time_sec)
+        while True:
+            await asyncio.sleep(sleep_time_sec)
+
+            try:
+                await job_store.cleanup_expired_jobs()
+                logger.debug("Expired jobs cleaned up")
+            except:  # noqa: E722
+                logger.exception("Error during job cleanup")
+
+    async def _submit_cleanup_task(self, scheduler_address: str, db_url: str, log_level: int = logging.INFO):
+        """Submit a cleanup task to the cluster to remove the job after expiry."""
+        logger.debug("Submitting periodic cleanup task to Dask cluster at %s", scheduler_address)
+        async with self.client(self._scheduler_address) as client:
+            self._periodic_cleanup_future = client.submit(self._periodic_cleanup,
+                                                          scheduler_address=self._scheduler_address,
+                                                          db_url=db_url,
+                                                          log_level=log_level)
+
+    @staticmethod
+    def _setup_worker():
+        """
+        Setup function to be run in each worker process. This moves each worker into it's own process group.
+        This fixes an issue where a Ctrl-C in the terminal sends a SIGINT to all workers, which then causes the
+        workers to exit before the main process can shutdown the cluster gracefully.
+        """
+        os.setsid()
+
+    async def run(self):
+
+        # Write the entire config to a temporary file
+        with tempfile.NamedTemporaryFile(mode="w", prefix="nat_config", suffix=".yml", delete=False) as config_file:
+
+            # Get as dict
+            config_dict = self.full_config.model_dump(mode="json", by_alias=True, round_trip=True)
+
+            # Three possible cases:
+            # 1. Dask is installed and scheduler_address is None, we create a LocalCluster
+            # 2. Dask is installed and scheduler_address is set, we use the existing cluster
+            # 3. Dask is not installed, we skip the cluster setup
+            dask_log_level = LOG_LEVELS.get(self.front_end_config.dask_log_level.upper(), logging.WARNING)
+            dask_logger = logging.getLogger("distributed")
+            dask_logger.setLevel(dask_log_level)
+
+            self._scheduler_address = self.front_end_config.scheduler_address
+            if self._scheduler_address is None:
+                try:
+
+                    from dask.distributed import LocalCluster
+
+                    self._cluster = LocalCluster(processes=True,
+                                                 silence_logs=dask_log_level,
+                                                 n_workers=self.front_end_config.max_running_async_jobs,
+                                                 threads_per_worker=1)
+
+                    self._scheduler_address = self._cluster.scheduler.address
+
+                    with self.blocking_client(self._scheduler_address) as client:
+                        # Client.run submits a function to be run on each worker
+                        client.run(self._setup_worker)
+
+                    logger.info("Created local Dask cluster with scheduler at %s", self._scheduler_address)
+
+                except ImportError:
+                    logger.warning("Dask is not installed, async execution and evaluation will not be available.")
+
+            if self._scheduler_address is not None:
+                # If we are here then either the user provided a scheduler address, or we created a LocalCluster
+
+                from nat.front_ends.fastapi.job_store import Base
+                from nat.front_ends.fastapi.job_store import get_db_engine
+
+                db_engine = get_db_engine(self.front_end_config.db_url, use_async=True)
+                async with db_engine.begin() as conn:
+                    await conn.run_sync(Base.metadata.create_all, checkfirst=True)  # create tables if they do not exist
+
+                # If self.front_end_config.db_url is None, then we need to get the actual url from the engine
+                db_url = str(db_engine.url)
+                await self._submit_cleanup_task(scheduler_address=self._scheduler_address,
+                                                db_url=db_url,
+                                                log_level=dask_log_level)
+
+                # Set environment variabls such that the worker subprocesses will know how to connect to dask and to
+                # the database
+                os.environ.update({
+                    "NAT_DASK_SCHEDULER_ADDRESS": self._scheduler_address,
+                    "NAT_JOB_STORE_DB_URL": db_url,
+                })
+
+            # Write to YAML file
+            yaml_dump(config_dict, config_file)
+
+            # Save the config file path for cleanup (required on Windows due to delete=False workaround)
+            config_file_name = config_file.name
+
+            # Set the config file in the environment
+            os.environ["NAT_CONFIG_FILE"] = str(config_file.name)
+
+            # Set the worker class in the environment
+            os.environ["NAT_FRONT_END_WORKER"] = self.get_worker_class_name()
+
+        try:
+            if not self.front_end_config.use_gunicorn:
+                import uvicorn
+
+                reload_excludes = ["./.*"]
+
+                # By default, Uvicorn uses "auto" event loop policy, which prefers `uvloop` if installed. However,
+                # uvloop’s event loop policy for macOS doesn’t provide a child watcher (which is needed for MCP server),
+                # so setting loop="asyncio" forces Uvicorn to use the standard event loop, which includes child-watcher
+                # support.
+                if sys.platform == "darwin" or sys.platform.startswith("linux"):
+                    # For macOS
+                    event_loop_policy = "asyncio"
+                else:
+                    # For non-macOS platforms
+                    event_loop_policy = "auto"
+
+                uvicorn.run("nat.front_ends.fastapi.main:get_app",
+                            host=self.front_end_config.host,
+                            port=self.front_end_config.port,
+                            workers=self.front_end_config.workers,
+                            reload=self.front_end_config.reload,
+                            factory=True,
+                            reload_excludes=reload_excludes,
+                            loop=event_loop_policy)
+
+            else:
+                app = get_app()
+
+                from gunicorn.app.wsgiapp import WSGIApplication
+
+                class StandaloneApplication(WSGIApplication):
+
+                    def __init__(self, app, options=None):
+                        self.options = options or {}
+                        self.app = app
+                        super().__init__()
+
+                    def load_config(self):
+                        config = {
+                            key: value
+                            for key, value in self.options.items() if key in self.cfg.settings and value is not None
+                        }
+                        for key, value in config.items():
+                            self.cfg.set(key.lower(), value)
+
+                    def load(self):
+                        return self.app
+
+                options = {
+                    "bind": f"{self.front_end_config.host}:{self.front_end_config.port}",
+                    "workers": self.front_end_config.workers,
+                    "worker_class": "uvicorn.workers.UvicornWorker",
+                }
+
+                StandaloneApplication(app, options=options).run()
+
+        finally:
+            logger.debug("Shutting down")
+            if self._periodic_cleanup_future is not None:
+                logger.info("Cancelling periodic cleanup task.")
+                # Use the scheduler address, because self._cluster is None if an external cluster is used
+                async with self.client(self._scheduler_address) as client:
+                    await client.cancel([self._periodic_cleanup_future], asynchronous=True, force=True)
+
+            if self._cluster is not None:
+                # Only shut down the cluster if we created it
+                logger.debug("Closing Local Dask cluster.")
+                self._cluster.close()
+
+            try:
+                os.remove(config_file_name)
+            except OSError as e:
+                logger.exception(f"Warning: Failed to delete temp file {config_file_name}: {e}")