I am currently trying to find the best way to keep two elastic clusters in sync. I tried it by fetching the modified data from the main cluster and update/insert to the backup cluster. The idea is to keep the two clusters in sync so that either of them can be main during failure. I am thinking taking snapshots from main and restoring them in the back cluster would do better job without fetching huge amount of modified data.
But I am not sure if two clusters can use the same snapshot repository to read and write data to.
Could you provide me insight into any better/best way to keep two elastic clusters in sync from a java application?