6a2481c6c6
Avoid keeping workflow in zombie state: on unexpected errors fail the workflow execution and save error information. If possible, fail the correspondent tasks, too. Done: - [x] Introduce state_info into Execution, fill it with error info when workflow fails. - [x] Add error handlers to `on_task_result` and `run_task` - [x] Add error handlers and tests to `start_workflow` - [x] Add tests for YAQL eval exceptions - [x] Add error handler and tests to `resume_worklfow` TODO: - [ ] Make _fail_workflow public and let user fail workflow via API (on separate commit) Closes-bug: 1415886 Change-Id: Ib1fd661580d3a426dbbc7401a3622a98c7840e00 |
||
---|---|---|
.. | ||
sqlalchemy | ||
v1 | ||
v2 | ||
__init__.py |