https://neovisionconsulting.blogspot.com/2020/01/pythonsanic1data.html
I am also researching here,
According to the blog above, even though Python can process 200,000 data per second using the Sanic framework, it is Go lang (Go language) that can really take advantage of the multicore of the CPU.
https://molpako.hatenablog.com/entry/2018/11/05/235541
Assuming that the content of the above blog is correct,
> Python 's GIL prevents multithreading from benefiting from multicore.
But,
In this blog, it seems that Python is multi-core and multi-thread compatible,
The real point is that you need to prepare a server with a spec that can perform a large-scale benchmark and a program for the website,
As of 2020.01.22.
Neither hardware nor software (program) could be prepared, so I just read and verified the blog and saw it.
Please understand.
In addition, the fastest Japronto framework in Python seems to be able to process 1 million DATAs per second. This is also the subject of research.
A large-scale web site that supports C10K problems without using AI and machine learning, etc.
If you take advantage of true multi-core & multi-thread support,
It is like Java or Go lang (Go language).
I would like to ask you to support true multi-core & multi-thread support in Python.
aon CEO
Python programmer
Masahiro Ishizuka
According to the blog above, even though Python can process 200,000 data per second using the Sanic framework, it is Go lang (Go language) that can really take advantage of the multicore of the CPU.
https://molpako.hatenablog.com/entry/2018/11/05/235541
Assuming that the content of the above blog is correct,
> Python 's GIL prevents multithreading from benefiting from multicore.
But,
Implement concurrency (multi-core, multi-thread) in Python
http://sfujimoto.hatenablog.com/entry/2017/01/06/030116In this blog, it seems that Python is multi-core and multi-thread compatible,
The real point is that you need to prepare a server with a spec that can perform a large-scale benchmark and a program for the website,
As of 2020.01.22.
Neither hardware nor software (program) could be prepared, so I just read and verified the blog and saw it.
Please understand.
In addition, the fastest Japronto framework in Python seems to be able to process 1 million DATAs per second. This is also the subject of research.
A large-scale web site that supports C10K problems without using AI and machine learning, etc.
If you take advantage of true multi-core & multi-thread support,
It is like Java or Go lang (Go language).
I would like to ask you to support true multi-core & multi-thread support in Python.
aon CEO
Python programmer
Masahiro Ishizuka
0 コメント:
コメントを投稿