Calling IndexManager.isIndexConsistent() while a foreground reindex is running may lead to corrupted index

Summary

Repeated calls to IndexManager.isIndexConsistent() while a foreground reindex is running may lead to corrupted index.

Note, that in Jira 7.x the API method that was responsible for determination of whether index is available, IndexManager.isIndexAvailable(), is defective. See JRASERVER-68900.

This problem was reported and reproduced on Jira 7.6, at the same time we couldn't reproduce it on Jira EAP 8. That being said, Jira 8 and corresponding Jira 7.x versions contain the proper fix.

Workaround

Currently there is no known workaround for this behavior. A workaround will be added here when available.