JIRA Indexing 7.8/7.13 + upgrade

[Karthik Ramachandran]

As part of our regular upgrade cycle (Minor/Patch), we redeploy JIRA in different server, and restore backup of DB to different server. Though as a Java application, we may just deploy the application alone, and not redeploy DB we have to follow the deployment principles within organization.

When we do the the same, JIRA doesn't start, and enforces re-indexing post DB restore. I'm looking at alternatives to reduce the indexing time as it creates big downtime. Below are some of the question that I don't have clarity. Any clarification on below would be helpful.

Indexing

1. JIRA indexes issue fields to provide better search experience. But, should I really need to re-index when I just restore the old copy of DB, and point to new JIRA server? I should be able to create directory copy and restore (or) create a snapshot and apply. In case of snapshot/apply, I'm sure if it is possible, as JIRA itself is not accessible without indexing. Restore page is available only after login, and navigate to Administration/System option.

2. Is my assumption as detailed earlier where I can just deploy the application with minor version upgrades and point to exiting DB/Storage work? Major versions may require follow an specific path, but not sure of minor versions

3. Does JIRA provide any API constructs that I can use to perform snapshot and restore? It would help automate the process

We are not using data center version and in single server mode. DC version may come in future.

Thanks for clarification.