We had been seeing clients with large datasets (1 M+), which have issues with re-indexing, even if there's no issues, re-index takes a lot of time, it is a big plus that we can leave the process on one node while we keep using the rest and cause no downtime, but sometimes it would be better to have a faster re-index, even for troubleshooting, waiting hours for a re-index to complete, or finding out that it failed at 70 - 90% progress, it is frustrating for users, admins and even support.

Please add the option to share re-index loads across the nodes servers, a load balancer that could divide work between nodes so their resources are also used to process the same re-index but processing different requests, this should allow system administrators to increase the index threads used, i.e. a 3 node data center Jira Software site with a default usage of 20 index threads (jira-config.properties), that shared between 3 nodes, I would like it to be 60 combined, (I'm not even sure if this is possible both from a developer and infrastructure side), but it should really be a BIGGER plus for the whole Atlassian Jira Data Center community.