* add a library which defines available datasources as constants
* include thread_pool gem and concurrent-ruby gem
* modify per process connection_pools global variable to be a concurrent map datastructure
* add a new DsConnectionPool libary to provide the api for using connections from a connection pool
* convert the connection pool parameters in the ENV to integers before using them
* add a new function in the ds_connection_pool library to reset_connection_pool/s
* add unit tests for the ds_connection_pool library
* fix typo
* better variable names and module_names
* refactor connection pool library test cases to use mocked query service
* add magic comment to make strings frozen
* remove airtable from the list of connection pooled datasources
* separate module inside available_data_sources to list connection poolable datasources
* use datasource_id as the key for storing connection pools of a type; Also handle cases when the pools connections are stale
* add additional test cases for the connection pool library
* fix rubocop errors
