Spaces:
Paused
Paused
File size: 11,719 Bytes
b9f94e1 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 | """Workflow engine adapters: Temporal, Prefect, Airflow."""
from __future__ import annotations
import logging
from abc import ABC, abstractmethod
from datetime import UTC, datetime
from typing import Any
from hermes.config.settings import get_settings
logger = logging.getLogger(__name__)
class WorkflowAdapter(ABC):
"""Abstract base for workflow engine adapters."""
@abstractmethod
async def connect(self) -> None:
"""Connect to the workflow engine."""
@abstractmethod
async def register_workflow(self, name: str, steps: list[dict[str, Any]]) -> None:
"""Register a workflow definition."""
@abstractmethod
async def execute_workflow(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Execute a workflow and return results."""
@abstractmethod
async def get_workflow_status(self, execution_id: str) -> str:
"""Get the status of a workflow execution."""
@abstractmethod
async def disconnect(self) -> None:
"""Disconnect from the workflow engine."""
class TemporalAdapter(WorkflowAdapter):
"""Temporal.io workflow engine adapter."""
def __init__(self) -> None:
self.settings = get_settings()
self._client: Any = None
async def connect(self) -> None:
"""Connect to Temporal server."""
try:
from temporalio.client import Client
self._client = await Client.connect(
self.settings.workflow.temporal_host,
namespace=self.settings.workflow.temporal_namespace,
)
logger.info(f"Connected to Temporal at {self.settings.workflow.temporal_host}")
except ImportError:
logger.warning("temporalio not installed. Run: pip install temporalio")
except Exception as e:
logger.warning(f"Failed to connect to Temporal: {e}")
async def register_workflow(self, name: str, steps: list[dict[str, Any]]) -> None:
"""Register workflow (Temporal handles this via decorators)."""
logger.info(f"Temporal workflow '{name}' registered with {len(steps)} steps")
async def execute_workflow(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Execute workflow on Temporal."""
result: dict[str, Any] = {"status": "pending", "execution_id": "", "output": {}}
if not self._client:
logger.warning("Temporal not connected, using local execution")
return await self._local_execute(name, context)
try:
handle = await self._client.start_workflow(
workflow=name,
arg=context or {},
id=f"{name}-{datetime.now(UTC).timestamp()}",
task_queue="hermes-tasks",
)
result["execution_id"] = handle.id
output = await handle.result()
result["output"] = output if isinstance(output, dict) else {"result": str(output)}
result["status"] = "completed"
except Exception as e:
logger.error(f"Temporal execution failed: {e}")
result["status"] = "failed"
result["error"] = str(e)
return result
async def get_workflow_status(self, execution_id: str) -> str:
"""Get workflow execution status from Temporal."""
if not self._client:
return "unknown"
try:
handle = self._client.get_workflow_handle(execution_id)
desc = await handle.describe()
status_map = {
1: "running",
2: "completed",
3: "failed",
4: "cancelled",
5: "terminated",
6: "continued_as_new",
7: "timed_out",
}
return status_map.get(desc.status.code, "unknown")
except Exception:
return "unknown"
async def _local_execute(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Fallback local execution when Temporal is unavailable."""
return {
"status": "completed",
"execution_id": f"local-{name}",
"output": context or {},
"note": "Executed locally (Temporal not connected)",
}
async def disconnect(self) -> None:
"""Disconnect from Temporal."""
self._client = None
logger.info("Disconnected from Temporal")
class PrefectAdapter(WorkflowAdapter):
"""Prefect workflow engine adapter."""
def __init__(self) -> None:
self.settings = get_settings()
self._client: Any = None
async def connect(self) -> None:
"""Connect to Prefect server."""
try:
from prefect.client import get_client
self._client = get_client()
logger.info("Connected to Prefect")
except ImportError:
logger.warning("prefect not installed. Run: pip install prefect")
except Exception as e:
logger.warning(f"Failed to connect to Prefect: {e}")
async def register_workflow(self, name: str, steps: list[dict[str, Any]]) -> None:
"""Register a Prefect flow."""
logger.info(f"Prefect flow '{name}' registered with {len(steps)} steps")
async def execute_workflow(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Execute a Prefect flow."""
result: dict[str, Any] = {"status": "pending", "execution_id": "", "output": {}}
if not self._client:
logger.warning("Prefect not connected, using local execution")
return await self._local_execute(name, context)
try:
from prefect.deployments import run_deployment
deployment_id = f"{name}/hermes-deployment"
flow_run = await run_deployment(
deployment_id,
parameters=context or {},
timeout=self.settings.workflow.timeout_seconds,
)
result["execution_id"] = str(flow_run.id)
result["output"] = flow_run.parameters or {}
result["status"] = "completed"
except Exception as e:
logger.error(f"Prefect execution failed: {e}")
result["status"] = "failed"
result["error"] = str(e)
return result
async def get_workflow_status(self, execution_id: str) -> str:
"""Get flow run status from Prefect."""
if not self._client:
return "unknown"
try:
from prefect.client import get_client
client = get_client()
flow_run = await client.read_flow_run(execution_id)
return flow_run.state.name.lower() if flow_run.state else "unknown"
except Exception:
return "unknown"
async def _local_execute(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Fallback local execution when Prefect is unavailable."""
return {
"status": "completed",
"execution_id": f"local-{name}",
"output": context or {},
"note": "Executed locally (Prefect not connected)",
}
async def disconnect(self) -> None:
"""Disconnect from Prefect."""
self._client = None
logger.info("Disconnected from Prefect")
class AirflowAdapter(WorkflowAdapter):
"""Apache Airflow workflow engine adapter."""
def __init__(self) -> None:
self.settings = get_settings()
self._client: Any = None
self._base_url = "http://localhost:8080"
async def connect(self) -> None:
"""Connect to Airflow API."""
try:
import httpx
airflow_user = self.settings.workflow.airflow_username or "airflow"
airflow_pass = self.settings.workflow.airflow_password or "airflow"
self._client = httpx.AsyncClient(
base_url=self._base_url,
auth=(airflow_user, airflow_pass),
timeout=30.0,
)
response = await self._client.get("/api/v1/health")
if response.status_code == 200:
logger.info(f"Connected to Airflow at {self._base_url}")
else:
logger.warning(f"Airflow health check failed: {response.status_code}")
self._client = None
except ImportError:
logger.warning("httpx not installed. Run: pip install httpx")
except Exception as e:
logger.warning(f"Failed to connect to Airflow: {e}")
self._client = None
async def register_workflow(self, name: str, steps: list[dict[str, Any]]) -> None:
"""Register an Airflow DAG."""
logger.info(f"Airflow DAG '{name}' registered with {len(steps)} steps")
async def execute_workflow(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Trigger an Airflow DAG run."""
result: dict[str, Any] = {"status": "pending", "execution_id": "", "output": {}}
if not self._client:
logger.warning("Airflow not connected, using local execution")
return await self._local_execute(name, context)
try:
response = await self._client.post(
f"/api/v1/dags/{name}/dagRuns",
json={
"conf": context or {},
"dag_run_id": f"hermes-{datetime.now(UTC).timestamp()}",
},
)
if response.status_code == 200:
data = response.json()
result["execution_id"] = data.get("dag_run_id", "")
result["status"] = "running"
else:
result["status"] = "failed"
result["error"] = f"Airflow API returned {response.status_code}"
except Exception as e:
logger.error(f"Airflow execution failed: {e}")
result["status"] = "failed"
result["error"] = str(e)
return result
async def get_workflow_status(self, execution_id: str) -> str:
"""Get DAG run status from Airflow."""
if not self._client:
return "unknown"
try:
response = await self._client.get(f"/api/v1/dagRuns/{execution_id}")
if response.status_code == 200:
return response.json().get("state", "unknown").lower()
except Exception:
pass
return "unknown"
async def _local_execute(self, name: str, context: dict[str, Any] | None = None) -> dict[str, Any]:
"""Fallback local execution when Airflow is unavailable."""
return {
"status": "completed",
"execution_id": f"local-{name}",
"output": context or {},
"note": "Executed locally (Airflow not connected)",
}
async def disconnect(self) -> None:
"""Disconnect from Airflow."""
if self._client:
await self._client.aclose()
self._client = None
logger.info("Disconnected from Airflow")
def get_workflow_adapter(engine: str = "") -> WorkflowAdapter:
"""Factory function to get the appropriate workflow adapter."""
if not engine:
settings = get_settings()
engine = settings.workflow.engine
adapter_map: dict[str, type[WorkflowAdapter]] = {
"temporal": TemporalAdapter,
"prefect": PrefectAdapter,
"airflow": AirflowAdapter,
}
adapter_cls = adapter_map.get(engine.lower())
if not adapter_cls:
logger.warning(f"Unknown workflow engine '{engine}', falling back to Temporal")
adapter_cls = TemporalAdapter
return adapter_cls()
|