aiometrics
Generate metrics from asyncio applications
How to install
As any other python package, use pip or include it in the install_requires section of your setup.py
pip install aiometrics
Sample
Sample application included below makes usage of the default StdoutDriver to print the metrics into the stdout.
# sample.py
import asyncio
import datetime
import random
import aiometrics
@aiometrics.trace
@asyncio.coroutine
def echo_delay():
ms = random.randrange(0, 5000)
t = ms/1000
yield from asyncio.sleep(t)
print(datetime.datetime.now())
@asyncio.coroutine
def main():
while True:
yield from echo_delay()
if __name__ == '__main__':
loop = asyncio.get_event_loop()
asyncio.ensure_future(main())
loop.run_forever()
loop.close()
Run this sample with the following command:
python sample.py
Here is the sample output:
2016-06-20 17:56:03.969607
...
2016-06-20 17:57:03.993648
{"instance": {"hostname": "dmo.local", "id": "acf9842e-c127-4844-8026-d7c518076ac2"}, "traces": {"__main__:echo_delay": {"2016-06-20T23:56:00+00": {"count": 24, "avg": 2510.3998333333334, "min": 229.113, "max": 4986.661}}}}
This generates a per-minute report listing metrics for every coroutine function decorated with @aiometrics.trace. Metrics include (per minute):
- Total number of executions
- Minimum execution time
- Maximum execution time
- Average execution time