Jump to: navigation, search

Difference between revisions of "TaskFlow/SaveAndRestore"

Line 4: Line 4:
  
 
What should be saved:<br />
 
What should be saved:<br />
1. Task name (or other identifier, unique for each task).<br />
+
1. Flow type id.<br />
2. Task data required to execute and revert.<br />
+
2. Flow kwargs.<br />
3. Task state (executed, failed, etc).<br />
+
3. Flow state.<br />
4. Executed actions ids.<br />
+
4. Task name (or other identifier, unique for each task).<br />
5. Flow type id.<br />
+
5. Task data required to execute and revert.<br />
6. Flow kwargs.<br />
+
6. Task state (executed, failed, etc).<br />
7. Flow state.<br />
+
7. Executed actions ids.<br />
  
 
Restoring the flow:<br />
 
Restoring the flow:<br />
 
'''Flow can be created from the blocks template (as a new flow)'''. Flow id will help to choose the correct template. Then Flow state, kwargs and tasks data should be restored and set. Executed actions should be added to the proper collections. After the Flow can be resumed.
 
'''Flow can be created from the blocks template (as a new flow)'''. Flow id will help to choose the correct template. Then Flow state, kwargs and tasks data should be restored and set. Executed actions should be added to the proper collections. After the Flow can be resumed.

Revision as of 14:59, 6 August 2013

Save and restore task flow. Brief description.

Flow must be saved each time when 'on_task_finished' and 'on_task_reverted' events are happened.

What should be saved:
1. Flow type id.
2. Flow kwargs.
3. Flow state.
4. Task name (or other identifier, unique for each task).
5. Task data required to execute and revert.
6. Task state (executed, failed, etc).
7. Executed actions ids.

Restoring the flow:
Flow can be created from the blocks template (as a new flow). Flow id will help to choose the correct template. Then Flow state, kwargs and tasks data should be restored and set. Executed actions should be added to the proper collections. After the Flow can be resumed.