In some situations JIRA DC have stale/lost cluster lock. Normally you can check this by navigating to Cluster Locks check [ System > Support Tools > Instance health > Cluster Locks ]. Problem that JIRA Admin space (mentioned check is under that space) uses lock com.atlassian.jira.config.DefaultReindexMessageManager.messageLock, so if it's lost, than it's impossible to check the locks.
Create separate Cluster Status space which:
- will not use any lock to access it
- will show all current Cluster locks with following data:
- Lock name - Node - Age
- (Advanced) will be able to break/steal Cluster lock
Check DB table clusterlockstatus: