Issue Summary

When creating issues in bulk where the assignee is being set via an automation rule that is using Round Robin assignment by group, defined by issue smart value, the issues are not being evenly distributed across all members of the group if the group returned from the smart value is not consistent for all issues that trigger the rule during the bulk operation. 

This behavior does not occur using the methods below:

• if issues are created in the UI natively 1 by 1
• if Round Robin assignment is based on "A user in a defined list" (Issues generated via REST API)
  

Steps to Reproduce

1. In Jira Administration > User management > Groups create multiple groups that will be associated with field values of issues input at issue creation.
2. Add multiple members to each group.
3. Create an automation rule to be triggered by issue creation. Add an action to assign the issue to a user in a group > Method of choose assignee should be set to "Round-robin" > and use a smart value to define which group to use when e valuating who to assign the issue to.

    i.e. {{issue.fields.customfield_10301.child.value}}

4. Create a set of issues using REST API using the /rest/api/2/issue/bulk endpoint. Example below where a .json file was prepared to be used in the curl command execution:

   curl -u username:password -X POST \
     -H "Content-Type: application/json" \
     --data @path/to/file/example_tickets.json \
     <baseURL>/rest/api/2/issue/bulk 

5. Review Issue assignment of issues that were created in bulk.

Expected Results

The issues that triggered the automation rule execution and assignment was set as a result should be evenly distributed across all members of each group

Actual Results

Issues are unevenly distributed across members of each group. This can be visualized using a "Two Dimensional Filter Statistics gadget where XAxis is set to Issue Type and YAxis is set to Assignee

Workaround

1. Adjust the rule configuration by adding a branch for each possible "group" value and then assign issue via Round Robin in each branch:
   
   • A drawback to this approach is Rule Execution duration will greatly increase due to each branch evaluation. You may consider increasing max.processing.time.per.day service limit if rule becomes throttled.
2. Create an individual rule for each possible group that will be returned during issue assignment. After the trigger condition you will want to add a JQL condition immediately afterward so that rule execution will only occur when the group value meets the requirement: