Attributes

Configuration options for the Durable Execution system

Attributes

The Durable Execution system is configured through the DurableExecution class, which provides the following attributes:

Attribute	Type	Default	Description
storage	DurableExecutionStorage	(required)	Storage backend for checkpoint persistence
execution_id	str	Auto-generated	Unique identifier for this execution
auto_cleanup	bool	True	Automatically cleanup checkpoints on completion
debug	bool	False	Enable debug logging

Configuration Example

from upsonic import Agent, Task, DurableExecution
from upsonic.durable import FileDurableStorage

# Create storage backend
storage = FileDurableStorage("./durable_checkpoints")

# Create durable execution
durable = DurableExecution(
    storage=storage,
    auto_cleanup=False,
    debug=True
)

# Create agent and task
agent = Agent("openai/gpt-4o")
task = Task(
    description="Process large dataset and generate report",
    durable_execution=durable
)

# Execute with checkpointing
result = agent.do(task)
print(f"Execution ID: {task.durable_execution_id}")
print(result)

# If execution fails, resume from checkpoint
# result = agent.continue_durable(
#     durable_execution_id=task.durable_execution_id,
#     storage=storage
# )