Summary

During the HA testing, we uncover some scenarios where business central repository can get corrupted if the pod is killed on the middle of an operation.
This JIRA cover this situations and some improvements recommended by engineering.

Goals
Add support for recovery scenarios (fail over)

• User is able to delete a corrupted project (caused by AF-2143)
• User is able to re-trigger an index in a project (fix for AF-2144)

Further improvements:

• Be able to know each pod user is logged in via UI on about screen (AF-2380)
• Alert on UI when ISPN or AMQ is down (AF-2381)
• Store last space opened in browser cookie (AF-2382)

Non goals

To fix other HA related issues that might be targeting further releases.
Make HA work under heavy load scenarios. BC authoring wasn't never designed for such.
Deal with unsaved data in a pod that was killed.

Acceptance criteria

User is able to delete a corrupted project.
User is able to re-trigger an index in a project.
User get notified when AMQ/ISPN is down
User are redirected to the last space where it worked on
All issues linked to this requirement are resolved.
QE verified that there is a regression coverage for fixed issues in community.