Yongwoon Cho

I am quite new to sqlalchemy and don't have experience about how heavy session create/commit/close operations are. I was just searching for the coding pattern for using tornado and sqlalchemy...

Would it be too bad if the framework calls commit functions every time query is called in thread pool? If it is insert query, I believe actual writing will happen...

I mean running session.commit in thread pool together with query function as this operation is expected to have IO operation. It can be done by adding a simple wrapper like...

``` import time import hashlib from absl import app, flags from tornado_sqlalchemy import SQLAlchemy from sqlalchemy import Column, Integer, String, Index db = SQLAlchemy() FLAGS = flags.FLAGS USER_PASSWORD_SALT = '12345678990'...

I added select query and measure time as well. ``` import time import hashlib from absl import app, flags from tornado_sqlalchemy import SQLAlchemy from sqlalchemy import Column, Integer, String, Index,...

I realized that it is not an easy problem. Because of lazy loading, every step can be blocking if model object is accessed. So, I think the whole block within...

Since the guide's indentation is 4, it is good to have 4-space indentation in all examples through out the document. There were 2-space examples which made me confused.