OpenSearch, instances vs. nodes

When it comes to Amazon OpenSearch Service pricing and instances, it's important to understand that instances and nodes are essentially the same thing in this context. Each instance you provision in OpenSearch Service represents a node in your cluster.

The pricing for OpenSearch Service is defined per instance, and there isn't a separate specification for node count because each instance is a node. Therefore, you don't need to consider nodes separately from instances when looking at pricing.

To address your specific question: 1 instance with 2 nodes is not a configuration that exists in OpenSearch Service. Instead, if you need two nodes, you would provision 2 instances with 1 node each. These two configurations are equivalent:

1. 2 instances with 1 node each
2. 2 nodes (which means 2 instances)

Both of these configurations would be priced the same, as you're essentially paying for two instances either way.

When considering Reserved Instances (RIs) for OpenSearch Service, you're making a commitment to use a specific instance type for a one or three-year term. The RI applies to the instance (node) itself, providing a discount compared to On-Demand pricing.

Remember that RIs are not flexible in terms of instance types. If you reserve a specific instance type, the discount only applies to that exact instance type. So, when planning your OpenSearch Service deployment and considering RIs, you should think in terms of the number and type of instances (nodes) you need, rather than trying to separate the concept of instances and nodes.
Sources
OpenSearch pricing question | AWS re:Post
Reserved Instances in Amazon OpenSearch Service - Amazon OpenSearch Service
Open-Source Search Engine - Amazon OpenSearch Service Pricing - AWS
Open-Source Search Engine - Amazon OpenSearch Service Features - AWS