Issue Summary

Auto-scheduling mechanism, in Advanced Roadmaps Jira plugin, sets wrong "Target start" or/and "Target end" dates for issues, which have child issues, but miss grand-child issue(s) at some level.

This is reproducible on Data Center: yes

Steps to Reproduce

Text version

1. Create a project of type Scrum, named “Test project 1”
2. Go to Configurations → Issues → “Issue types”
3. Create new issue type of standard type and name it “Iteration” (the name was chosen arbitrary)
4. Create another new issue type of standard and name it “Requirement” (the name was chosen arbitrary)
5. Go to Configurations → Issues → “Issue type schemes”
6. Click on “Edit” next to the “TP1: Scrum Issue Type Scheme” project scheme.
7. Move issue types “Task”, “Bug” and “Sub-task” from column “Issue Types for Current Scheme” (left) to column “Available Issue Types” (right)
8. Move issue types “Iteration” and “Requirement” from column “Available Issue Types” (right) to column “Issue Types for Current Scheme” (left)
9. Click on “Save” button.
10. Go to Configurations → “Manage apps” → “Hierarchy configuration” (to the left, in “ADVANCED ROADMAPS FOR JIRA” section).
11. Create new level called “Requirement” and select “Requirement” issue type as the only issue type for this level.
12. Create another new level called “Iteration” and select “Iteration” issue type as the only issue type for this level.
13. Make sure, that level “Requirement” is located on top of “Iteration” level.
14. Click on “Save changes”.
15. Create new plan named “Test plan 1” with “Test project 1” as the only issue source of type project.
16. In (newly created) plan view, go to “Teams” tab.
17. Click on “…” to the only team, that is diplayed on the screen. Then click on “Edit”.
18. Change the type of the team from “Kanban” to “Scrum” and add “admin” user as a single team member.
19. Click on “Save”.
20. Go to plan configurations → Scheduling and set “Estimate issues in” to “Story points”, then uncheck “Use sprint dates when issues don't have start and end dates” checkbox.
21. Go back to the plan view → Teams tab.
22. Click on “…” on the only team, that is displayed on the screen. Then click on “Edit”.
23. Set “Velocity (pts)” to 10 and “Iteration length (weeks)” to 2.
24. Click on “Save”.
25. Go to plan view → Backlog tab.
26. Set hierarchy from “Requirement” to “Story”.
27. Click on “Fields” and add “Story points” field to the view.
28. Create 2 issues of type Requirement and call them “Req 1” and “Req 2” respectively.
29. Under “Req 1”, create 2 issues of type Iteration and call them “Iter 1” and “Iter 2” respectively.
30. Under “Req 2”, create 2 issues of type Iteraction and call them “Iter 3” and “Iter 4” respectively.
31. Under “Iter 1” create 2 issues of type Epic and call them “Epic 1” and “Epic 2” respectively.
32. Under “Epic 2” create 1 issue of type Story and call it “Story 1”.
33. Under “Iter 2” create 2 issues of type Epic and call them “Epic 3” and “Epic 4” respectively.
34. Under “Epic 3” create 1 issue of type Story and call it “Story 2”
35. Under “Epic 4” create 1 issue of type Story and call it “Story 3”
36. For “Story 3”, “Story 2”, “Story 1” and “Epic 1” set “Story points” to 6.
37. Click on “Review changes” and then “Save selected changes to Jira”.
38. Now click on “Auto-schedule” and then “Preview results”.
39. Observe the wrong end date for “Req 1”. It it now equal to the end date of the “Epic 1” (and “Iter 1”), although it should be equal to the end date of the “Story 1” (and “Epic 2” and “Iter 2”). See Screenshot 1, section in red rectangle.
40. Observer the correct start and end date for “Req 2”. They are equal to the start date of “Story 2” and the end date of “Story 3” respectively. See Screenshot 1, section in green rectangle.

Screenshot 1

Video version

https://www.loom.com/share/3f18af85ee674fa4ac8f59d86db0ca86

Expected Results

Issues, which have child issues, should get "Target start" set to the "Target start" of the earliest child issue and "Target end" date set to the "Target end" date of the latest child issue (see Screenshot 1, section in green rectangle)

Actual Results

Issue, which have child issues, but have incomplete child branches ("unbalanced" tree of issues), receive bad "Target start" and/or "Target end" dates. Some child issue(s) might get missed while searching for the earliest "Target start" and/or latest "Target end" dates because of missing child issues in their case. 

Workaround

A workaround, in this case, is to make sure, that the "tree of issues" is complete, meaning that if a particular issue's child issue has its own child issues, all its sibling issues should also have child issues to make the tree of issues complete.