CloudSearch, a new product by Amazon Web Services (AWS), offers integration of highly scalable and fully managed search functionality in their software applications. It uses the same A9 technology which runs Amazon.com’s search. In order to use the new functionality, IT staff needs to create a search domain as well as upload the data which they need searchable. CloudSearch will the provision automatically the resources needed as well as the indexes required. In order to enable data searches, the data must be in the Search Data Format which can be accomplished through XML text files or JavaScript Object Notation (JSON).
When the searchable data grows in number, CloudSearch has the capability to scale it automatically. Users can also have the capability to transform search parameters, use new setting, and fine tune the relevance of the search without any need for more data upload. The AWS Management Console is used to change settings and administer other Amazon’s services. With the new search feature, Amazon will continue to attract those customers who do not have much IT experience and expertise.
Billing for the said service, however, is not that simple. Customers are billed monthly for data transfer, IndexDocuments requests, document batch uploads, and search instances. CloudSearch has the capability to determine the search instances and size in order to offer high throughput and low latency search performance. It creates an index and gets the right initial search instance in order to make sure that the index is storable in RAM. Instance type can cost $0.68, $0.48, and $0.12 per hour depending on its size. Changes in the configuration of the index through IndexDocuments can cost $0.98 per GB of data in the search domain. $0.10 is also charged per thousand of requests of batch upload which can be up to a maximum of 5MB. CloudSearch also charges when data is transferred out of it. In the East Coast, $0.12 per GB is charged for the first 10 TB.