elasticsearch-river-couchdb

ES isn't really meant to be used as an operational datastore (although I appreciate some do use it that way). CouchDB provides strong durability and master-master replication. It's also very easy to sync changes from CouchDB (https://github.com/elasticsearch/elasticsearch-river-couchdb) to ES so they make a good complimentary pair. Use CouchDB for the raw data (where you want strong durability and write scalability) and use ES to search over it.

I currently run a setup where I interact with CouchDB and have ElasticSearch use the couchdb river plugin [0] to pretty much replicate an exact copy of the data for ad-hoc queries. There have been times when I thought about just interacting with ES directly rather than going through the CouchDB layer. I'm still not 100% sure if I should or not. I don't need any consistency guarantees for most of my data and I'm ok with a few records going missing in case of a crash. I like CouchDB, but I have to use the CouchBase single server dev preview to get snappy compression and the copy on write approach to data storage makes the on-disk files grow pretty large over time. Compaction currently takes several hours :-/ [0] https://github.com/elasticsearch/elasticsearch-river-couchdb...