Looking for a independent 'ancient history mark' to track a ES index for Mongo access

I am considering using elastic search to act as an index into my mongo collections. But it can take some time for ES to get the most recent data resulting in race conditions that are problematic. Anyone with an opinion regarding a good way to implement an 'ancient history mark' to help avoid looking at stale es index. My whole env is on AWS.

I also have a rabbit MQ message bus. I was thinking AHM could be a good subject that all interested hosts could subscribe.

Just at the start of my investigation and have not committed to this option. Other options could be Redis Memcache or even a tiny SQL table but the whole point was to stay performant.