Expanding the Cloud - Introducing Amazon CloudSearch
Today Amazon Web Services is introducing Amazon CloudSearch, a new web service that brings the power of the Amazon.com's search technology to every developer. Amazon CloudSearch provides a fully-featured search engine that is easy to manage and scale.
Context & Ripple Effects
This is the next step in a familiar AWS sequence: the platform began by renting out single pieces of Amazon.com's own infrastructure — storage with S3 in 2007, then the SimpleDB database that same December — before broadening compute itself, including Windows Server workloads on EC2 in 2008. CloudSearch extends the same playbook to search, turning technology built for Amazon.com's retail site into a metered web service.
The timing matters competitively: six months earlier Google made an explicit run at AWS with new storage and App Engine offerings, and AWS is widely regarded as the largest cloud provider even without independent verification. Productizing another internal capability both widens the service gap and deepens the lock-in question raised by the August 2011 EC2 outage that briefly took down several high-traffic sites.
First-order effects
- Developers can now buy search as a managed API instead of operating their own Lucene/Solr-style clusters, moving indexing, scaling, and tuning costs onto AWS's bill.
- For AWS, CloudSearch is another line item that makes the platform stickier: a customer who builds site search on it faces higher switching costs than one using only S3 or EC2.
Second-order effects
- Hosted-search vendors and consultancies selling custom search deployment lose their easiest customers — small and mid-size teams who previously had no cheaper managed option.
- Google, which framed its October 2011 storage and App Engine push as a direct challenge to AWS, is pressed to answer with a comparable managed-search offering of its own rather than ceding another service tier.
Third-order effects
- The structural pattern is web-scale companies externalizing internal systems until every layer of the application stack becomes a commodity, pay-per-use service — search following storage, database, and compute down the same path.
- Each new managed tier concentrates more operational dependency on one vendor, so reliability events like the 2011 EC2 outage carry escalating systemic weight across the sites built on these services.
The trend: AWS keeps converting Amazon.com's internal infrastructure into metered public services, expanding the platform faster than rivals can match service-for-service.