Amazon Web Services has released its cloud search offering, CloudSearch, which has been rumored for months. The new service is based on the same technology that powers product searches on Amazon.com. Amazon is the first company to offer search-as-a-service, but something tells us the Internet giant’s competitors will soon be scrambling to keep up now that Amazon has entered the market.
Search, Amazon Style
Amazon thoroughly dominates the public cloud service market, providing massive amounts of computing capacity and long list of services. Now they’ve added a fully managed search service to the list of things they provide, moving Amazon Web Services (AWS) even closer to being a one-stop-shop for all things cloud.
Amazon is no stranger to search. Its retail site offers sophisticated product search features, and Amazon formed child company A9 to focus on search in 2003. Amazon has a huge amount of search knowledge and massive volumes of structured and unstructured data stored on AWS. Search was a natural next step, and the service will likely be widely embraced by AWS customers.
Many casual observers will likely compare Amazon’s CloudSearch to Google’s enterprise search offering. However, the products aren’t really equivalent.
Google provides a search appliance that provides a plug-and-search experience for customers that don’t need a significant amount of tuning or customization. The device also lives on the customer’s network. In contrast, Amazon’s search service is on the public cloud and behaves more like a traditional search solution that must be loaded with data and can be highly customized to return results in a very specific way.
CloudSearch is really more similar to Lucid Imagination’s Lucene/Solr-based cloud offering or the many enterprise search-as-a-service offerings like Searchperience and Searchify.
Although Amazon’s new search service is attractive, it’s unlikely to cause companies to dump their internal search platforms en masse. No matter how fast or easy to configure the platform is, it’s still in the public cloud, which is a big no for many organizations with sensitive internal data. However, for companies without security concerns, CloudSearch should definitely be on the list of platforms to consider if search functionality is required in their application and site.
Search, no matter what tool you use, has always been notoriously complex. CloudSearch simplifies the infrastructure part of the equation by providing on-demand scalability and processing power. Amazon touts that users can get started with CloudSearch in three steps and under an hour:
- Create a search domain, which is really just a data container.
- Upload the data to be searched. In search vernacular, each record is called a document and must be formatted in a specific way for CloudSearch to index it.
- Perform searches.
In reality, a bit more work is required, as described in the following video.
CloudSearch is priced similarly to other AWS services. However, potential users should note that they have no control over size instance used for their search data. AWS determines that based on the size of data and index necessary for supporting searches.