Lithops-1.2.0

Added

• New local_executor() to run pywren jobs in the local machine
• New localhost compute backend
• New localhost storage backend
• New docker_executor() to run pywren jobs in the local machine by using docker
• New docker compute backend

Changed

• Docs updated
• Code refactor

Fixed

• Internal fixes
• Bump pillow from 5.4.1 to 6.2.0

Lithops-1.1.1

Added

• Allowed partitioner to split files by a number of chunks
• Missing logic in knative backend

Changed

• Docs updated

Fixed

• Internal fixes

Lithops-1.1.0

Added

• Added knative-serving compute backend
• Added Dockerfile skeleton for slim Python3.6 runtime (only 307MB)
• Added CACHE_DIR in ~/.pywren/cache
• knative_executor() and function_executor()
• support to work on multiple regions at a time

Changed

• Docs updated
• Runtime Dockerfiles updated
• Runtime requirements updated
• Updated Cloudpickle lib to version 1.2.2
• Parameters introduced in the executor now overwrite the config
• updated tests

Fixed

• Internal logic to generate runtime_metadata
• invalid call to "is_remote_cluster" method
• Cloudpickle lib to accept any kind of import
• include_modules option in serializer

Lithops-1.0.20

Added

• Storage abstraction for data partitioner
• Added 'extra_params' arg to map() and map_reduce() calls
• Logic to reuse IAM API Key tokens during 1 hour
• More debug logging

Changed

• Docs updated
• Full support for Python3.5

Fixed

• Fixed minor issue in config
• Fixed possible issue extracting metadata from large docker images (runtimes)

Lithops-1.0.19

Added

• Added 'obj' as an optional arg for the functions when a user wants to process objects from OS
• Added 'rabbitmq' as an optional arg for the functions
• Added 'id' as an optional arg for the functions
• Added rabbitmq example

Changed

• Deleted 'bucket' 'key' 'data_stream' function args in favor of 'obj'
• Internal improvements related data partitioning
• Changed create_timeline_plots() method name to create_execution_plots()
• Docs updated
• updated notebooks
• Upgrade cos-sdk Python module version

Fixed

• Fixed tests
• Fixed CVE-2019-12855 security alert

Lithops-1.0.18

Added

• Added CloudObject abstraction
• Added CloudObject example
• Restored OOM exception
• Allowed to get_results when rabbit monitoring is activated
• Allowed rabbimq to monitor multiple jobs at a time
• Statuses returned from rabbitmq to futures

Changed

• Code refactoring about compute abstraction
• Reorganized libs folder
• Updated cloudpickle lib from 0.6.1 to 1.2.1
• Updated glob2 lib to 0.7
• Updated tests
• Modified job_id format

Fixed

• Fixed minor issue listing CF actions
• Fixed issue when executing pywren inside pywren
• Fixed possible issue with invalid config parameters
• Fixed wrong method name: build_runtime()
• Fixed internal_storage parameter in partitioner
• Fixed crete_timeline_plots method according recent changes

Lithops-1.0.17

Changed

• Code refactoring about compute abstraction

Fixed

• Fixed issue with invocation retries

Lithops-1.0.16

Added

• Added missing init file
• Allowed 'clean=all' arg in clean() method

Changed

• Simplified invoker
• Moved compute and storage classes to separate files
• Deleted unnecessary files
• Close plots on finish
• Code refactoring about compute abstraction

Fixed

• Fix in invocation method

Lithops-1.0.15

Added

• Added log information
• Added init files
• Store runtime_metadata into a local cache in order to reduce exec time

Changed

• Modularized Invoker
• Changed some variable names
• Docs updated

Fixed

• Fixed set_memory in invoker
• Fixed unneeded memory usage
• Fixed none finished futures
• Fixed wait method
• Fixed issue with map_reduce()

Lithops-1.0.14

Added

• Pywren runtime deployment as script
• Changed name of the runtime deployment script
• Added 'pywren_runtime clean' option
• Added function_name and runtime_memory to future's status
• Added 'pywren_runtime update all' option
• Added exception when preinstalled modules list is not well provided
• Add package parameter to delete function

Changed

• Improved sending execution statuses through rabbitmq
• Improved exception management
• Moved some logs to debug
• Improved runtime deployment script
• Changed logic order of the create_timeline_plots method

Fixed

• Preventing false out-of-memory error
• Fixed issue when using rabbitmq to monitor executions
• Fixed issue tracking map_reduce execution with progressbar
• Some other fixes