I've ran into this problem in several single-node installations: Free disk space became less than 15%, and ElasticSearch crashed. Upon clearing the space and restarting, I could not access the indices. Had to reallocate them on the node in order for server state to change from red to yellow.
Can I prevent this situation?
For example, is there any way for me to set a limit in which indexing of new items is rejected, yet enough space is left for ElasticSearch server to function correctly until the problem is resolved?