Lithops

Added
- [Core] New "warm_container" and "func_result_size" in future statistics
- [Core] New logic to detect referenced modules/libs

Changed
- [Core] New monitoring system
- [Core] Deleted strong dependency to pika==0.13.1

Fixed
- [Partitioner] Fixed partitioner when obj url contains more than one subfolder
- [Cli] Fixed serverless runtime lifecycle methods

Deleted
- [Core] Removed cloudpickle from lithops.libs

Added
- [Core] Allow Support for Python 3.9
- [Core] Added standalone get_result() and wait() methods
- [knative] Include GCP and Azure storage libs into default knative runtime
- [CodeEngine] Enable internal kubernetes pod executions
- [k8s] Enable internal kubernetes pod executions
- [Cli] Added 'empty' function to storage cli to empty a bucket
- [Core] Added new method to FunctionExecutor() to calculate execution costs
- [IBM CF] Added formula to calculate execution costs
- [Multiprocessing] Added Nanomsg connection type for addressable backends
- [Multiprocessing] Added expiry time for Redis multiprocessing resources
- [Multiprocessing] Added Listener and Client for multiprocessing using Redis
- [Azure Functions] Added support for http trigger
- [Core] Set lithops to localhost mode when config files is empty

Changed
- [IBM CF] Change user_key to API-key pass instead of user
- [Azure] Changed configuration keys
- [Core] Improved worker when chunksize is set to values > 1
- [Core] Check lithops version mismatch in host instead of in worker

Fixed
- [Core] Overwrite the runtime set in config with the runtime set in the FunctionExecutor
- [Cli] Fixed --config param in lithops cli
- [Standalone] Fixed internal executions
- [Core] Fixed rabbitmq monitor when get_result() is called after wait()
- [GCP Storage] Fix GCP Storage backend put obj as stream
- [GCP Functions] Improved runtime create time
- [Azure blob] Fix in azure blob get method
- [Azure Functions] Fix build runtime command

Added
- [Core] Added multiprocessing support in workers
- [Core] Added 'cunksize' param to API calls
- [Core] Added 'worker_processes' param to API calls
- [Core] Allow a worker to process more than one call trough 'chunksize' param
- [Core] All Functions logs are now synchronized with the client
- [Config] Allow 'log_level' and 'log_format' keys in configuration
- [Config] Allow 'log_stream' and 'log_filename' keys in configuration
- [Config] Allow 'runtime' being configured at serverless backend level
- [Config] Allow 'invoke_pool_threads' being configured at serverless backend level
- [Multiprocessing] Added generic Manager
- [Kubernetes] Add kubernetes job backend
- [CLI] Extended lithops cli with storage put, get, delete and list options
- [Azure] Added missing azure functions backend methods

Changed
- [Core] Improved Standalone execution mode
- [Core] Renamed utils.setup_logger() method to utils.setup_lithops_logger()
- [Core] Renamed partitioner 'chunk_size' param to 'obj_chunk_size'
- [Core] Renamed partitioner 'chunk_n' param to 'obj_chunk_number'
- [GCP Cloud Run] Refactor backend, removed 'gcloud' CLI calls.
- [IBM VPC] Improved IBM VPC backend
- [AWS Lambda] Lambda layer modules update

Fixed
- [Multiprocessing] Fix issues related to Pipes and Queues
- [Multiprocessing] Fix multiprocessing.context methods
- [CodeEngine/knative] Fix getting docker username in MAC OS hosts

Fixed
- [Code Engine] Fixing code engine docker image

Added
- [Joblib] Joblib backend upgraded
- [AWS Lambda] Support for container-based runtimes
- [AWS Lambda] Support for running functions in a VPC
- [AWS Lambda] Support for attaching EFS volumes
- [Core] Added cloudpickle, tblib and ps-mem deps as requirement of the runtimes
- [Core] Add a new Serverless mode that allows to include the function within the runtime

Changed
- [Core] Allow Standalone mode to start 1 VM per activation

Fixed
- [Core] Fixed issue in clean when it is called between wait and get_result
- [Core] Fixed multiprocessing Queue and get_context
- [Core] Fixed multiprocessing args mapping in map and map_async
- [Localhost] Fixed issue when using docker images in Windows or MAC

Deleted
- [Core] Removed tblib from lithops.libs
- [Core] Removed ps-mem from lithops.libs

Required actions
- Run: python3 -m pip install -U cloudpickle tblib ps-mem ibm-vpc namegenerator

Added
- [Azure] Azure Functions backend upgraded
- [Azure] Azure blob backend upgraded
- [Alibaba] Alibaba Functions backend upgraded
- [Alibaba] Alibaba Storage backend upgraded
- [Localhost] Support passing file-like objects to put_object
- [Localhost] Support head_bucket and head_object storage operations

Changed
- [Core] Moved tests.py script to 'scripts' folder

Fixed
- [Core] Fixed Storage API error when no config is provided
- [Core] Fixed expired IAM token in IBM CF during an execution
- [Core] Minor fixes in multiprocessing API
- [Core] Fixed executor logging
- [Core] Fixed issue in cleaner between wait and get_result
- [CodeEngine] Fixed issue getting region
- [Localhost] Fixed empty parent directory deletion when deleting objects
- [Localhost] Made list_keys/list_objects behavior consistent with other backends