changes.MD

Changelog for each release

Changelog for 0.7.11

• Issues resolved in 0.7.10 milestone: https://github.com/radical-cybertools/radical.entk/milestone/12
• Separated the CU creation from RMQ Communication thread

Changelog for 0.7.9

• Issues resolved in 0.7.9 milestone: https://github.com/radical-cybertools/radical.entk/milestone/11?closed=1
• Included ability to suspend and resume pipeline execution
• Several bug fixes

Changelog for 0.7.8

• Hotfix for issue 259

Changelog for 0.7.7

• Issues as part of 0.7.7 milestone: https://github.com/radical-cybertools/radical.entk/milestone/10/
• Improved test coverage

Changelog for 0.7.6

• Minor fixes to be able to upload to conda repo

Changelog for 0.7.5

• Issues as part of 0.7.5 milestone: https://github.com/radical-cybertools/radical.entk/milestone/9/
• Documentation improved with more examples
• Bug fixes

Changelog for 0.7.4

• Bug fixes
• Improved documentation
• Improved test coverage

Changelog for 0.7.3

• Bug fixes
• Improved documentation
• Improved test coverage

Changelog for 0.7.0

• Issues as part of 0.7.0 milestone: https://github.com/radical-cybertools/radical.entk/milestone/8/
• API Changes:
  • 'cores' attribute of task changed to cpu_reqs and gpu_reqs which are dictionaries with the following structure:

    task.cpu_reqs =     {
                            'processes': 1,
                            'process_type': None/MPI,
                            'threads_per_process': 1,
                            'thread_type': None/OpenMP
                        }
    
    task.gpu_reqs =     {
                            'processes': 1,
                            'process_type': None/MPI,
                            'threads_per_process': 1,
                            'thread_type': None/OpenMP
                        }

    • ResourceManager object is not exposed to the user. The resource description is to be provided to the AppManager using the 'resource_desc' attribute

    amgr = AppManager()
    amgr.resource_desc = {
        'resource': 'local.localhost',
        'walltime': 10,
        'cpus': 1}

    • AppManager does not have assign_workflow() method. Instead you assign the workflow using the assignment operator (similar to resource desc) to the workflow attribute.

    amgr = AppManager()
    amgr.workflow = pipelines

    Note: pipelines can be a list or a set of Pipeline objects but there are no guarantees of order
    • AppManager has two important additional arguments:
      • write_workflow (True/False) to write the executed workflow to a file post-termination
      • rmq_cleanup (True/False) to cleanup the rabbitmq queues post-execution
    • AppManager reads default values from a JSON config file. Expected config file structure:

    {
    "hostname": "localhost",
    "port": 5672,
    "reattempts": 3,
    "resubmit_failed": false,
    "autoterminate": true,
    "write_workflow": false,
    "rts": "radical.pilot",
    "pending_qs": 1,
    "completed_qs": 1,
    "rmq_cleanup": true
    }

    • The shared_data attribute is part of the AppManager object now (since the resource manager is not exposed to the user anymore).

    amgr = AppManager()
    amgr.shared_data = ['file1.txt','/tmp/file2.txt']

Changelog for 0.6.3

• Added some more tests

Changelog for 0.6.2

• Several issues addressed
• Bulk submission of tasks across entire set of pipelines
• Write workflow structure post-execution upon 'write_workflow=True' argument to AppManager

Changelog for 0.6.1

• Multiple bug fixes

Changelog for 0.6.0

• _parent_pipeline and _parent_stage on Stage and Task objects changed to parent_pipeline and parent_stage.