Work in progress
Motivations:
- While working on python based web applications, we end up tuning the processes and treads model based on the infrastructure and performance requirements. If we have any connection caching strategy in place, the tuning become complex since we are dealing with process/threads + connection cache. Having a connection pool proxy helps with managing the connection cache strategy independent of the process/threads model.
- Connection caching at the application level would be local to a Python VM process and a connection established in a process can not be shared with other processes.