Amazon Technologies, Inc.

David John Ward, Jr

Stephen Alden Elliott

David Nunnerley

Brian Helfrich

Peng Zhai

Matthew Scott Garman

Dhanvi Harsha Kapila

Rachid Ennaji

Derek Solomon Pai

Geoffrey Scott Pare

Jin Dong Kim

Kiran Kumar Meduri

Michael R Foster

Christopher Brian Barclay

Gautam Shanbhag

Mohammed Samir Alabsi

Gregory Branchek Roth

Marvin Michael Theimer

Peter F Hill

Don Johnson

Peter Sven Vosshall

Arun Sundaram

Kurt Kufeld

Peter N De Santis

Laura Ellen Grit

Léon Thrane

Bradley D Roberts

Dennis Pilarinos

Adam K Loghry

Karthigeyan Seetharaman

Jin Seop Kim

John W Gray, Iii

Rusenia Carpio Ingram

Gregory Evan Redman

Fran McClellan Shigley, Iii

Methods and apparatus for dynamic modification of interruptibility settings of network-accessible resources are disclosed. A system includes a resource manager and a plurality of resource instances, each of which has an interruptibility setting. In response to an instance acquisition request, the resource manager allocates an instance with a first interruptibility setting, allowing the resource manager to revoke the client's access to the instance without a notification. In response to an approval of an interruptibility upgrade request, the resource manager modifies the interruptibility setting to a second setting that allows the client to retain access to the instance for at least a specified time. Respective billing amounts for the client's use of the instance are determined for each of the interruptibility settings used.

Dynamic modification of interruptibility settings for network-accessible resources

In a multi-tenant web services platform, a request is received from a first customer requesting that a second customer be allowed to use a computing resource associated with the first customer. Control of the computing resource is transferred to the second customer based on a constraint selected by the first customer. A first portion of a cost associated with the computing resource is billed to the first customer. The control of the computing resource by the second customer is terminated in accordance with the constraint selected by the first customer.

Lendable computing resources

Methods and apparatus for dynamic pricing and scheduling of network data transfers are disclosed. A system a plurality of network-accessible resources and a data transfer coordinator. In response to a client request for a network transfer of a data set, the coordinator determines, based at on a network-related metric of one or more resources, at least one time window during which a portion or all of the data set is transferable for a pricing rate below a threshold. The coordinator notifies the client of the time window and provides a duration-limited token authorizing data transfer during the window. The coordinator determines whether a network transmission of data on behalf of the client is valid based at least in part on a time at which the transmission is initiated and at least in part on a use of the token.

Dynamic scheduling for network data transfers

Methods and apparatus for flexible capacity pool reservations for network-accessible resources are disclosed. A system includes a plurality of resources of a provider network, and a resource manager. The resource manager receives a reservation modification request from a client that has an existing reservation for an original resource capacity and an original reservation duration. The modification request specifies a desired resource capacity. The reservation manager identifies resources to provide the desired capacity, and determines pricing components for the requested modification based on the difference between the desires resource capacity and the original capacity and on the remaining portion of the duration of the existing reservation. The resource manager modifies the reservation as requested and notifies the client regarding the pricing components.

Flexible capacity reservations for network-accessible resources

A provider network may implement resource management and billing for logical and physical availability zones. Consolidated billing techniques may be applied to the aggregate usage of multiple linked user accounts. A billing setting may indicate whether consolidated billing is performed according to resource consumption in logical availability zones or physical availability zones. Mappings to physical availability zones may be provided to linked user accounts so that computing resources may be consumed according to either logical or physical availability zones. Capacity reservations for computing resources for a particular physical availability zone may be made available to each of the linked user accounts for consumption.

Resource management for logical and physical availability zones of a provider network

A technology is described for prioritizing a plurality of interruptible computing instances. An example method may include receiving a request to interrupt an executing interruptible computing instance that may be included in a group of interruptible computing instances having assigned instance priority values. The instance priority values may establish an order in which the interruptible computing instances may be interrupted. The interruptible computing instance may be identified that has a lower instance priority value as compared to other instance priority values assigned to the interruptible computing instances. The interruptible computing instance identified may then be terminated.

Interruptible computing instance prioritization

A resource manager may monitor resources offered by a resource provider. The resource manager receives computing metrics from the resources and may perform a management action when a customer defined usage limit is satisfied. The management action may include the resource manager generating an alert, or the resource manager preventing the customer from further accessing the resources offered by the resource provider. The resource manager may also be used to reduce the cost of running a customer's application that uses resources offered by the service provider. The resource manager may also receive business metrics from the customer and determine rates of return using the business metrics and the computing metrics.

Resource manager

One or more computing instances are instantiated and allocated to customers. The computing instances have a modifiable interruptibility property that is operable to allow a computing instance to be terminated or reallocated from a first customer to a second customer. Requests for additional computing instances having modifiable interruptibility properties are received. Information pertaining to the instantiated computing instances and requested additional computing instances is provided. The information may include a current status of the instantiated computing instances or requested additional computing resources.

Notification of computing capacity status

Methods and apparatus for managing user-defined capacity reservation pools for network-accessible resources are disclosed. A system includes a plurality of resource instances of a provider network, and a resource manager. In response to a pool assignment request from a client specifying a pool identifier of a reserved instance pool created on behalf of the client, the resource manager assigns one or more resource instance reservations of the client to a pool with the first pool identifier. In response to an instance activation request from the client, the resource manager identifies a particular pool created on behalf of the client from which to select a particular resource instance reservation for activation, and activates the particular resource instance.

User-defined capacity reservation pools for network-accessible resources

Methods and apparatus for notification-based pricing of excess cloud capacity are disclosed. A system includes a resource manager and a plurality of resource instances, each of which has an interruptibility setting. In response to a first acquisition request, the resource manager allocates a first instance with a first interruptibility setting to a first client, allowing the resource manager to revoke the first client's access to the first instance without a notification. In response to an approval of a second acquisition request, the resource manager allocates a second instance with a second interruptibility setting to a second client, allowing the client to retain access to the instance for a delay interval after the second client receives an access revocation notification. Respective billing amounts for the clients' use of the instances are determined based on the interruptibility settings used.

Notification based pricing of excess cloud capacity

Growing community of inventors

Seattle, WA, United States of America

David John Ward, Jr