Amazon Technologies, Inc.

Ryan Farris

Benjamin Aldouby Schwartz

Ronil Sudhir Mokashi

Prashant Verma

Saurav Sengupta

Akhil Ramachandran

Meera Jindal

Michael Vidyadhar Kale

Abilash Ravikumar

Ahmet Gurcan

Khawaja Salman Shams

Andrew Kent Warfield

David Alexander Dunlap

Mathew Persons Jack

Mark DeSpain

Jimmy George

Jesse Robert Hepburn

Marcin Manek

Dylan Kyle Abraham

Prakash Bhasker

Matei Mitaru Berceanu

Colleen Marie Toth

In some embodiments, a system is provided, and computer-executable instructions cause the system to: obtain, at an edge server of a content delivery network, a request for a content item from a client device; determine that the there are multiple versions of the first content item that require varying amount of resources to serve including a first version, and a second version requiring more resources to serve than the first version; select the second version; send the second version; determine that the client device is not properly rendering the second version of the content item; and in response to determining that that the client device is not properly rendering the second version, send the first version.

Automatically serving different versions of content responsive to client device rendering errors

A set of data units associated with a data object is obtained, such that respective instances of the data object can be reconstructed from respective subsets of the set. Corresponding to a request for the data object, a first subset of the set is identified. The first subset meets a uniqueness criterion with respect to other subsets of the set that are used to respond to other requests for the data object. An instance of the data object is reconstructed from the first subset.

Space-efficient techniques for generating unique instances of data objects

A media storage service may store a plurality of copies of a same media recording in cloud DVR, one copy specific to one of a plurality of clients. The client may interact directly with the media storage service based on path-based indexing techniques for playback of the media recording. The client may send a request, including a path of a manifest file associated with the copy specific to the client, to the media storage service. The manifest file may include information indicating respective paths of one or more segments of the copy. The media storage service may identify and return the manifest file to the client. The media storage service may further receive requests from the client to access the segments of the copy. The requests may include the respective paths, based on which the media storage service may retrieve the segments for the client to play.

Using path-based indexing to access media recordings stored in a media storage service

This disclosure describes techniques for distributed storage of files that represent video content across different storage tiers to reduce the amount of computing resources used to store the files, while maintaining a low latency when responding to user requests to stream the video content. Video streaming services support the streaming of video content at different resolutions to support various user devices and/or user preferences. To provide streaming of a particular video content item at different resolutions, the video streaming services may store multiple files that represent the video content item for the different resolutions and stream at different bitrates to user devices. The techniques described herein include storing files configured to stream a video content item at different bitrates (and for different display resolutions) in a tiered storage structure to reduce the amount of computing resources taken to store the files.

Distributed storage of files for video content

A media storage service may maintain an asset including information to indicate a specific media content and a specific number of replicas to be stored by the media storage service for the specific media content. The media storage service may receive content recording-related requests from a recording manager to define or update the specific number for the specific media content. The media storage service may receive one or more segments specific to the media content from an encoder. The media storage service may create and store the specific number of replicas, as indicated in the asset, for the media content.

Asset-level management of media recording in cloud DVR systems

In some embodiments, a system is provided, and computer-executable instructions cause the system to: receive, at an edge server of a content delivery network (CDN), a request for a first video and a request for a second video; determine that the first video is not cached and that the second video is cached; request the first video from the origin server and log a cache miss; obtain the first video from the origin server and send responsive to the request; send the second video responsive to the request and log a cache hit; obtain a metric indicative of the volume of cache misses, including for the first and second videos, across edge servers of the CDN; and determine, based on the metric, an amount by which to scale resources implementing the origin.

Automatically augmenting user resources dedicated to serving content to a content delivery network

A content delivery network ('CDN') is provided herein that predicts content resources (e.g., a data object, such as a video file, an audio file, a script, an image, a document, etc.) that may be requested by a user device in the future and transmits or pushes such resources to the user device prior to receiving a request. The CDN may use artificial intelligence models, such as Markov models, in order to predict which content resources to retrieve and transmit proactively to the user device. The predictive techniques implemented by the CDN may reduce a latency of delivering requested content resources and/or a latency of the user device in rendering and displaying a content page.

Predictive content push-enabled content delivery network

Growing community of inventors

Bellevue, WA, United States of America

Ryan Farris