Amazon Technologies, Inc.

Shawn Jones

Eric Samuel Stone

Derek Solomon Pai

Ahmed Usman Khalid

Rishi Baldawa

Marcel Robert Guzman

Norman Jordan

Kunal Chopra

Mingzhe Zhu

Benjamin Jeffery

Arijit Choudhury

Alexander Kulikov

Siva Swaroop Palli

Felipe De Aguiar Kamakura

Mukti Ranjan Sahoo

Mihir Rajendrabhai Patel

Nikita Pinski

Igor Subbotin

Gustavo Soto Ridd

Asim Himani

A system for redistributing traffic in response to failure of a service region. An event bus service is configured to monitor performance of computing resources in a first region of the plurality of regions, wherein events are received from one or more event sources at the event bus via a network endpoint created for the event bus and delivered to one or more event targets by the event bus service in the primary region. Based on the monitoring of the performance of the computing resources in the first region, the event bus service detects a region change event for the event bus in the first region, after detection of the region change event, redirects further events to the event bus in the secondary region, wherein the further events are delivered to the one or more event targets by the event bus service in the secondary region.

Network endpoint with failure recovery and traffic redistribution

A system is configured to replicate events to be processed by service regions. The system implements an event bus service that hosts an event bus in respective ones of a plurality of regions. The event bus service is configured to receive, at the event bus in a secondary region of the plurality of regions, a replicated indication of an event that is replicated from the event bus in a primary region of the plurality of regions. The event bus service determines, at the event bus in the secondary region, whether the primary region has completed processing of the event. Based on a determination that the primary region has not completed processing of the event within the pre-determined time period, the event bus service processes the event according to the replicated indication and indicates, in the replication record, that the secondary region has completed processing of the event.

Replicated events in a multi-region network

Techniques are described for enabling a cloud-based service to provide rate controls for events to be delivered by the service to event targets. An event bus service, for example, broadly enables users to create and configure event buses that receive events from event sources and from which the event bus service filters, transforms, routes, and delivers events to selected event targets according to configurable rules. A new event egress call pattern is provided that estimates a rate of attempted event traffic to event targets and a rate of successfully delivered event traffic (also referred to as a local rate estimation), and that uses the rate estimations to determine when it may be appropriate to throttle event delivery, thereby reducing stress on downstream systems experiencing capacity issues and potentially reducing a number of event delivery retries to be handled by the event bus service.

Egress rate control in an event bus service

A system is configured to simulate outages of network resources. The system is configured to provide a control plane for computing resources of a provider network. The control plane is configured to cause simulated outages of a primary region of the plurality of regions selected to host the plurality of different computing resources. During the simulated outages, the control plane moves respective workloads of the plurality of different computing resources to be performed in the one or more secondary networks and tracks a performance of the one or more secondary regions hosting the moved respective workloads of the plurality of computing resources. After completing individual ones of the simulated outages of the first network, the control plane moves the respective workloads of the plurality of different computing resources back to the primary region.

Simulated network outages to manage backup network scaling

A computing resource service provider may provide computing instances organized into logical groups, such as auto-scaling groups. Computing instances assigned to an auto-scaling group may be place into standby. Standby instances may still be managed by the auto-scaling group but may not contribute to the capacity of the auto-scaling group for auto-scaling purposes.

Standby instances for auto-scaling groups

A computing resource service provider may provide computing instances organized into logical groups, such as auto-scale groups. Computing instances assigned to an auto-scale group may be associated with one or more load balancers configured to direct traffic to the computing instances. Furthermore, customers of the computing resource service provider may add or remove load balancer from the auto-scale groups. A background process may be used to add and remove computer instances of the auto-scale group from the load balancers customers are attempting to have added or removed.

Background processes in update load balancers of an auto scaling group

A transaction directed to a hierarchy of data objects stored in a non-transactional data store may be executed. The transaction may be received and description of the transaction stored in a parent data object of the hierarchy. Upon storing the description of the transaction, the transaction may be acknowledged as committed. Access requests for the hierarchy of data objects may be serviced by accessing the description of the transaction and applying to child data objects in the hierarchy returned for servicing a query directed to the hierarchy. A new transaction that is received may initiate application of the transaction according to the description in the parent data object with respect to child data objects.

Executing transactions for a hierarchy of data objects stored in a non-transactional data store

A secondary index may be implemented for a distributed data set that is strongly consistent. Updates to a distributed data set that add or remove items from the distributed data set may be reflected in the secondary index as part of performing the update. Pointers to items to be added to a distributed data set may be included in the secondary index as part of processing an insertion request for the new items. Pointers to items removed from a distributed data set may be removed from the secondary index as part of processing a deletion request. Changes to the secondary index may be performed so that the secondary index does not fail to identify items that are present in the distributed data set.

Strongly-consistent secondary index for a distributed data set

Migration of mutable data sets between data stores may be implemented. Multiple data sets may be stored at a source data store. Some of the data sets may be identified for migration to a destination data store. Updates to the identified data sets may be blocked while the identified data sets are copied from the source data store to the destination data store. In some embodiments, the copies of the data sets at the destination data store are validated. Updates to the data sets may then be allowed at the destination data store instead of the source data store.

Migration of mutable data sets between data stores

A computing resource service provider may provide computing instances organized into logical groups, such as auto-scale groups. Computing instances assigned to an auto-scale group may be associated with one or more load balancers configured to direct traffic to the computing instances. Furthermore, customers of the computing resource service provider may add or remove load balancer from the auto-scale groups. A background process may be used to add and remove computer instances of the auto-scale group form the load balancers customers are attempting to have added or removed.

A computing resource service provider may provide computing instances organized in to logical groups, such as auto-scaling groups. Computing instances assigned to an auto-scaling group may be detached from auto-scaling group. Detached instances may no longer be managed by the auto-scaling group and may not contribute to the capacity of the auto-scaling group for auto-scaling purposes.

Detaching instances from auto-scaling group

A computing resource service provider may provide computing instances organized into logical groups, such as auto-scale groups. Computing instances assigned to an auto-scale group may be associated with one or more load balancers configured to direct traffic to the computing instances. Furthermore, customers of the computing resource service provider may add or remove load balancer from the auto-scale groups.

Managing load balancers associated with auto-scaling groups

A computing resource service provider may provide computing instances organized into logical groups, such as auto-scaling groups. Computing instances assigned to an auto-scaling group may be associated with one or more load balancers configured to direct traffic to the computing instances. Furthermore, customers of the computing resource service provider may add or remove load balancer from the auto-scaling groups. The computing resource service provider may track a state of the load balancer in a database in order to enable efficient addition and removal of load balancers to auto-scale groups.

Growing community of inventors

Vancouver, Canada

Shawn Jones