Files
hatchet/examples/python/blocked_async/worker.py
Gabe Ruttner 8e80faf2d6 Fe overhaul docs (#1640)
* api changes

* doc changes

* move docs

* generated

* generate

* pkg

* backmerge main

* revert to main

* revert main

* race?

* remove go tests
2025-04-30 14:10:09 -07:00

49 lines
1.3 KiB
Python

import hashlib
import time
from datetime import timedelta
from hatchet_sdk import Context, EmptyModel, Hatchet
hatchet = Hatchet(debug=True)
# WARNING: this is an example of what NOT to do
# This workflow is intentionally blocking the main thread
# and will block the worker from processing other workflows
#
# You do not want to run long sync functions in an async def function
blocked_worker_workflow = hatchet.workflow(name="Blocked")
@blocked_worker_workflow.task(execution_timeout=timedelta(seconds=11), retries=3)
async def step1(input: EmptyModel, ctx: Context) -> dict[str, str | int | float]:
print("Executing step1")
# CPU-bound task: Calculate a large number of SHA-256 hashes
start_time = time.time()
iterations = 10_000_000
for i in range(iterations):
hashlib.sha256(f"data{i}".encode()).hexdigest()
end_time = time.time()
execution_time = end_time - start_time
print(f"Completed {iterations} hash calculations in {execution_time:.2f} seconds")
return {
"step1": "step1",
"iterations": iterations,
"execution_time": execution_time,
}
def main() -> None:
worker = hatchet.worker(
"blocked-worker", slots=3, workflows=[blocked_worker_workflow]
)
worker.start()
if __name__ == "__main__":
main()