The Redlock algorithm suggested for use with Redis has been the subject of some of criticism: https://martin.kleppmann.com/2016/02/08/how-to-do-distributed-loc

The Redlock algorithm suggested for use with Redis has been the subject of some of criticism: https://martin.kleppmann.com/2016/02/08/how-to-do-distributed-locking.html https://martin.kleppmann.com/2016/02/08/how-to-do-distribute... Kleppman is the author of Designing Data-Intensive Applications and generally someone whose opinion I trust in such things. I love Redis and use it extensively at $dayjob, but I stick to Consul for managing distributed locks. Consul was built from the ground up to handle such things; Redis handles it as a bit more of an afterthought / consequence of other features.