SIENTIAPDE-1350: Implement scheduled cleanup workflow using Temporal schedules. Adds schedule creation to worker startup and configures environment variables.

model_manager/schedules/cleanup_schedule.py

@@ -0,0 +1,85 @@
+"""Schedule configuration for cleanup workflow."""
+
+import os
+from datetime import timedelta
+
+from sientia_do.observability.logger import Logger as SientiaLogger
+from temporalio.client import (
+    Client,
+    Schedule,
+    ScheduleActionStartWorkflow,
+    ScheduleSpec,
+)
+
+# Schedule configuration from environment variables
+SCHEDULE_ID = os.getenv('CLEANUP_SCHEDULE_ID', 'cleanup-files-daily')
+CLEANUP_CRON = os.getenv('CLEANUP_CRON', '0 0 * * *')  # Default: midnight UTC
+CLEANUP_TIMEZONE = os.getenv('CLEANUP_TIMEZONE', 'UTC')
+CLEANUP_TASK_QUEUE = os.getenv('CLEANUP_TASK_QUEUE', 'cleanup_queue')
+CLEANUP_EXECUTION_TIMEOUT_HOURS = int(os.getenv('CLEANUP_EXECUTION_TIMEOUT_HOURS', '1'))
+
+
+async def schedule_exists(
+    client: Client, schedule_id: str, logger: SientiaLogger, metadata: dict[str, str | None]
+) -> bool:
+    """
+    Check if a schedule already exists.
+
+    Args:
+        client: Temporal client instance
+        schedule_id: ID of the schedule to check
+
+    Returns:
+        True if schedule exists, False otherwise
+    """
+    try:
+        async for schedule in await client.list_schedules():
+            if schedule.id == schedule_id:
+                return True
+        return False
+    except Exception as e:  # noqa: BLE001
+        logger.custom_error(f'Error checking if schedule exists: {e}', metadata)
+        return False
+
+
+async def create_cleanup_schedule(
+    client: Client, logger: SientiaLogger, metadata: dict[str, str | None]
+) -> None:
+    """
+    Create or update the cleanup files schedule.
+
+    This function is idempotent and can be called multiple times safely.
+    It will only create the schedule if it doesn't already exist.
+
+    Args:
+        client: Temporal client instance
+    """
+    # Check if schedule already exists
+    if await schedule_exists(client, SCHEDULE_ID, logger, metadata):
+        logger.custom_info(
+            f"Schedule '{SCHEDULE_ID}' already configured, skipping creation", metadata
+        )
+        return
+
+    await client.create_schedule(
+        SCHEDULE_ID,
+        Schedule(
+            action=ScheduleActionStartWorkflow(
+                'cleanup_files',
+                {},  # Empty input, will use default bucket from environment
+                id=f'cleanup-files-scheduled-{SCHEDULE_ID}',
+                task_queue=CLEANUP_TASK_QUEUE,
+                execution_timeout=timedelta(hours=CLEANUP_EXECUTION_TIMEOUT_HOURS),
+            ),
+            spec=ScheduleSpec(
+                cron_expressions=[CLEANUP_CRON],
+                time_zone_name=CLEANUP_TIMEZONE,
+            ),
+        ),
+    )
+
+    logger.custom_info(
+        f"Schedule '{SCHEDULE_ID}' created successfully. "
+        f'Cleanup will run at: {CLEANUP_CRON} ({CLEANUP_TIMEZONE})',
+        metadata,
+    )

model_manager/worker/worker.py

@@ -37,6 +37,7 @@ with workflow.unsafe.imports_passed_through():
 
     from model_manager import metrics
     from model_manager.activities.activities import Activities
+    from model_manager.schedules.cleanup_schedule import create_cleanup_schedule
     from model_manager.utils.connectors_config import (
         build_minio_config,
         build_mlflow_config,
@@ -117,6 +118,14 @@ async def main():
 
     logger.custom_info(f'Temporal client initialized at {host}', metadata)
 
+    # Create cleanup schedule (idempotent - only creates if doesn't exist)
+    try:
+        await create_cleanup_schedule(temporal_client, logger, metadata)
+    except Exception as e:  # noqa: BLE001
+        logger.custom_error(f'Failed to configure cleanup schedule: {e}', metadata)
+        # Don't fail the worker startup if schedule creation fails
+        # The schedule can be created manually if needed
+
     workers = [
         Worker(
             temporal_client,