Amazon Technologies, Inc.

Ram Cherukuri

William Wallace Allocca

James Cade Bodley

Nandan Ulhas Telang

Nitin Sivakrishnan

Harsha Ramalingam

Guy Bialostocki

Bharathi Masilamani

Chad Howard Young

Daniel James Hermanson

Daniel Lee Phelps

Ram Gole

Anthony James Wilson

Karl Clifford Bell

Atif Choudhry Rafiq

John Alexander DiSanti

The systems and techniques described herein optimize processing of requests, such as external requests, internal requests or inter-service requests. In some examples, a service may interact with a client device and determine, based at least in part on a state of the interaction with the client device, a potential client request that may subsequently be received from the client device. The service may then perform one or more actions to at least partially fulfill the potential client request.

Optimization of production systems

The techniques described herein provide software testing of a candidate version of software. In some examples, an interceptor may perform an asynchronous and non-blocking intercept of at least one request to a production system and may provide the intercepted at least one request to a shadow proxy service as at least one shadow request. The shadow proxy service may process the shadow request by causing the at least one shadow request to be classified based at least one criteria. The processed shadow requests are logged. A selection may be made from among the logged shadow request to be replayed by the production system based on one or more criteria used to classify the shadow requests. The shadow proxy service may replay the selected at least one logged shadow request to the production system.

A normalized caching system is configured to modify one or more parameters of a service request to generate a 'normalized' service request, causing the service to return a normalized response. The normalized response is typically useful for servicing a wider range of subsequent requests than is a non-normalized response. Thus, a wider range of subsequent service requests can be serviced from the cache, without the need to send such service requests to the service.

Parameter normalization for improved cache hit rate

A community of authors, readers and contributors interact with each other through a communication framework. The authors share creative works, such as fictional stories, for consumption by the readers and for commenting by the contributors. The readers may see successive drafts of the work as an author completes and revises the work. The contributors may provide comments or other contributions to assist the author. For example, the contributors may make suggestions about how to change the work or provide supplemental content, such as cover artwork, to the author. The author may recognize publicly contributors who provide helpful comments. The likelihood of an author completing a final draft of a work may be determined based on the author's activities regarding the work. The likelihood of a completed work being popular with readers may also be determined based on user interaction with a draft of the work.

Story development and sharing architecture

The techniques described herein provide software testing of a candidate version of software. In some examples, an interceptor intercepts at least one production request to a production version of the software and issues the production request to a shadow proxy service as a shadow request. The shadow proxy service causes the at least one shadow request to be processed by the candidate version of the software being validated and an authority version of the software being used to validate the candidate version. The shadow proxy service may then compare and/or analyze at least one candidate response to the shadow request from the candidate version and at least one authority response to the shadow request from the authority version. A dashboard service may provide at least some of the resulting information and issue a request the shadow proxy service to replay at least one of the shadow requests.

Shadow test replay service

Story development and sharing architecture: predictive data

A normalized caching system may receive a data request from a requestor, the request associated with a requested service. The normalized caching system may generate a normalized request by modifying one or more parameters of the request. A normalized response is typically useful for servicing a wider range of subsequent requests than is a non-normalized response. The system may then check whether an entry is cached for the normalized request. If no cached entry is found, the normalized request may be submitted to the requested service, and response received in return. The normalized response may be cached using the normalized request. The response may be denormalized in order to generate a response which is provided to the requestor. When a subsequent request is received, if that subsequent request normalizes to the same normalized request, it may be fulfilled using the response stored in cache, rather than using the requested service.

Growing community of inventors

Bellevue, WA, United States of America

Ram Cherukuri