Lithops

Added
- [AWS S3] Allow to use a public bucket without credentials
- [IBM] Automatically login to the container registry if the credentials are present in config
- [IBM] Force --platform=linux/amd64 in the "lithops runtime build" command
- [k8s] Added boto3 as a dependency of the default runtime
- [IBM VPC] Automatically get the ubuntu image ID
- [IBM VPC] Allowed to reuse floating IPs
- [IBM VPC] Automatically create resources if not provided in config
- [IBM VPC] Added 'region' config parameter
- [Partitioner] Allow wildcards in the object reference

Changed
- [IBM VPC] Reduced the number of mandatory config parameters
- [IBM VPC] Renamed profile_name config param to worker_profile_name
- [IBM VPC] Renamed ip_address config param to floating_ip

Fixed
- [AWS EC2] Fix spot instance price
- [Cli] Fix wrong config in the "lithops runtime update" command
- [Standalone] Fix missing private IP address
- [VM] Fix VM standalone backend

Added
- [Config] Allow to store the config file in "/etc/lithops/config"
- [CLI] Allow to specify 'memory' and 'version' in the 'lithops runtime delete' command
- [GCP Cloudrun] Allow setting min_workers to the autoscaler
- [GCP Functions] Added https trigger
- [Function Executor] Added additional arguments to pass to the reduce function in map_reduce()
- [AWS] Added session token as optional

Changed
- [Core] Include function result in the status.json file if it is < 8KB
- [Core] Include python version in the lithops worker name

Fixed
- [Serverless] Allow to delete runtimes from different lithops versions
- [AWS Batch] Fixed list_runtimes() method
- [localhost] Fixed localhost storage backend head method

Added
- [Google Cloud Functions] Added Python 3.10 runtime compatibility
- [Core] Allow to automatically transfer .so (cythonized .py) files if referenced in the code

Changed
- [Core] Improved cython coverage
- [IBM VPC] Make 'image_id' mandatory in config
- [IBM VPC] Infer zone_name from subnet
- [Knative] Reduced service name length
- [AWS EC2] Updated ec2 default ubuntu image to 22.04

Fixed
- [IBM VPC] Create floating IP using the specified resource group
- [IBM VPC] Attach floating IP to the subnet
- [Multiprocessing] Fix 'cpu_count' function from multiprocessing API
- [Code Engine] Add CE conflict exception to retriables
- [Core] Show logs from module_dependency.py
- [GCP Functions] Fix runtime_build command
- [Infinispan] Fix Infinispan storage backend
- [Core] Detect a Class if passed as a lithops input function

New
- [Azure Container APPs] Added new 'Azure Container APPs' serverless backend

Added
- [Azure Container APPs] Added new lithops backend for Azure Container APPs
- [Knative] Added Kourier as the networking layer
- [AWS Lambda] Added "env_vars" and "ephemeral_storage" parameters for lambda runtime config
- [Azure Functions] Allow to build a runtime from a custom requirements.txt file
- [K8s] Append 'docker_server' as a prefix of the runtime
- [Code Engine] Append 'docker_server' as a prefix of the runtime
- [Knative] Append 'docker_server' as a prefix of the runtime
- [Google Cloud Storage] Add missing logic in gcp_storage
- [Google Cloud] project_name and service_account are no longer required in config
- [Google Cloud] Allow to use GOOGLE_APPLICATION_CREDENTIALS for service_credentials
- [Google Cloud Run] Allow CPU values <1, 6 and 8
- [Alibaba Cloud Functions] Added Python 3.9 runtime compatibility
- [Alibaba Cloud Functions] Allow to build a runtime from a custom requirements.txt file
- [Core] Add support for building container runtimes with podman
- [Core] Delete only runtimes from the specified backend on 'lithops clean'
- [Executor] Added obj_newline parameter in map() and map_reduce() methods
- [Infinispan] Support DIGEST authentication by default with the Infinispan REST backend

Changed
- [Core] Load docker command only when needed instead of loading it always
- [Core] Load runtime data only on invocation
- [Google Cloud] project_name and service_account config parameters are no longer required
- [Multiprocessing] Improved remote logging
- [Monitor] Reduce debug log verbosity, status is printed every 30s or when a future changes state
- [AWS Batch] Increased resource limits
- [Executor] Changed 'reducer_one_per_object' parameter to 'obj_reduce_by_key'

Fixed
- [Google Cloud Functions] Fixed errors when deploying a function
- [Core] Make sure all temp files generated during build_runtime() are cleaned
- [Core] Fix internal partitioner
- [knative] AttributeError: 'KnativeServingBackend' object has no attribute 'coreV1Api'
- [knative] Fixed service deployment
- [Alibaba Cloud Functions] Fixed errors when deploying a function
- [Azure Functions] Fixed errors when deploying a function
- [Azure Functions] Fixed issue that was preventing building runtimes from a non-Linux system
- [Code Engine] Fix runtime_timeout not being applied

Added
- [Code Engine] Improved except-retry logic
- [IBM COS] Enables separate IAM authorization for COS and Compute backends

Added
- [Core] Add support for Python 3.10
- [Storage] Added `download_file()` and `upload_file()` methods to Storage API to enable multipart upload/download
- [AWS Lambda] Added 'architecture' option in config to allow deploy arm64 runtimes
- [AWS Batch] Added 'service_role' config parameter
- [IBM VPC] add gpu support for ibm_vpc backend
- [Standalone] Added docker login to standalone setup script
- [AWS EC2] Automatically get the default Ubuntu 20.04 AMI when not present in config
- [Code Engine] Added retry logic on connection exception

Changed
- [CLI] Renamed "lithops runtime create" command to "lithops runtime deploy"

Fixed
- [AWS Lambda] Fixed "layer builder" function creation necessary to deploy the default runtime
- [AWS Lambda] Allow to create container runtimes whose names contain '.', '/' and '_'.
- [AWS Lambda] List only the runtimes deployed by the current user
- [AWS Lambda] Do not raise an exception if the runtime is already deployed
- [Standalone] Fix cloudinit initialization script
- [Future] Fix host_status_query_count stat
- [Google Cloud Run] Fixed wrong variable name 'runtime_cpus'
- [Google Cloud] Changed docs for Google cloud backend regarding to id instead of name