In this benchmark I tested sync and async servers

There are 4 files in the benchers package

• async_bench.py
• sync_bench.py

These two are the main benchmark files. They send requests to their appropriate server. The number of requests they send is in the __init__.py file of this package. Also they log everything happens in the benchmark.

• async_sample.py
• sync_sample.py

These two files are just really small requesters for small and command line results. You can see their results with

python sync_sample.py

There are 3 pictures in the repo.

• plot_whole.png
• async_plot.png
• sync_plot.png

In this image, I plotted the results of both async (orange) and sync (blue) requesters. There is no doubt that the sync version takes really longer than the async version.

In these two pictures you can see the actual times against the number of requests which are done by that requester.

To produce the results yourself you have to do the following:

clone the repo

git clone https://github.com/mahdihaghverdi/grpc-bench

go inside

cd grpc-bench

create a virtual environment

python -m venv .venv

install the requirements

pip install -r requirements.txt

activate the venv (linux)

source .venv/bin/activate

• async test

run the server

python servers/async_server.py

in another terminal, run the bencher

python benchers/async_bench.py

• sync test

python servers/sync_server.py

python benchers/sync_bench.py

to plot the results

• whole

python plot_whole.py

• async

python async_plot.py

• sync

python sync_plot.py

to see and try the sample benchmarks

• async

python benchers/async_sample.py 

• sync

python benchers/sync_sample.py 

>>> import os
>>> os.uname()
posix.uname_result(sysname='Linux', nodename='mahdi', release='5.19.0-46-generic', version='#47~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jun 21 15:35:31 UTC 2', machine='x86_64')
>>>

>>> import platform
>>> platform.python_version()
'3.10.6'
>>> platform.python_build()
('main', 'May 29 2023 11:10:38')
>>> platform.python_compiler()
'GCC 11.3.0'
>>> platform.python_implementation()
'CPython'
>>>