IndexNotFoundException can be seen when the portal is initiated with a database populated with data from a previous hash

• Type: Bug
• Status: Closed
• Resolution: No Longer Reproducible
• Affects Version/s: Master
• Fix Version/s: None
• Component/s: Workflow
• Labels: None
• Fix Priority: 4

Current issue: IndexNotFoundException can be seen when the portal is initiated with a database populated with data from a previous hash

Steps to reproduce:

1. Checkout liferay-portal repo to 9091b47b798c8c08603abbddeb3949602fcb0784
2. Build portal (ant all)
3. Start the portal with a brand new MySQL database (lportal)
4. Associate Single Approver with blogs entry
5. Create few entries and approve some of them and go to Metrics to be sure the data is there
6. Go to CP > Search > Field Mappings and observe that workflow metrics indexes are not following the pattern: [<prefix>]workflow-metrics-instances<companyId> (LPS-108828)
7. Stop the portal
8. Rebase liferay-portal with latest master hash
9. Build portal (ant all)
10. Start the portal with the same database used on step 3
11. Go to CP > Metrics
12. Go to CP > Search Field Mappings

Actual result: IndexNotFoundException can be seen. Workflow metrics indexes persist without the correct pattern

Expected result: No exception should be seen and workflow metrics indexes should be displayed with the pattern described on LPS-108828

Environment:
master DXP old hash: 9091b47b798c8c08603abbddeb3949602fcb0784
master DXP new hash: 43f034138129c4601c1dd0cdf076d8cfc4e99c01