Describe correct opensearch usage

Co-authored-by: Morten Lied Johansen <[email protected]>
Co-authored-by: Kyrre Havik <[email protected]>

Author: 3 people

docs/persistence/opensearch/README.md

@@ -8,6 +8,15 @@ tags: [opensearch, explanation, persistence, services]
 
 OpenSearch is a fork of Elasticsearch that is maintained by Amazon. It is a drop-in replacement for Elasticsearch, and is fully compatible with the Elasticsearch API. It is a community-driven project that is open source and free to use.
 
-OpenSearch is a distributed, RESTful search and analytics engine capable of solving a growing number of use cases. It is a good choice for storing data that is not relational in nature.
+OpenSearch is a distributed, RESTful search and analytics engine capable of solving search related use cases. It is a good choice for implementing search over documents.
 
-NAIS offers OpenSearch via [Aiven](https://aiven.io/). Aiven OpenSearch can be used by applications in all environments, but must be *defined* in a GCP cluster.
+OpenSearch can index documents without storing them allowing you to save on storage space at the cost of not being able to retrieve the contents of the document.
+This enables patterns where you use OpenSearch as a way of indexing into other forms of persistence.
+Combining opensearch with a separate master data store such as PostgreSQL is a very common pattern.
+Using PosgreSQL with OpenSearch in this way gives you the same data durability and backup policies as you have with PostgreSQL.
+
+A very powerful pattern for OpenSearch is search and filter, also known as [faceted search](https://opensearch.org/docs/latest/aggregations/bucket/index/).
+
+NAIS offers OpenSearch via [Aiven](https://aiven.io/).
+Aiven OpenSearch can be used by applications in all environments, but must be *defined* in a GCP cluster.
+All OpenSearch instances on Aiven, regardless of plan, are backed up but there are no disaster backups outside of the provider.