Amazon Technologies, Inc.

Ryan Charles Schmitt

Bryan James Donlan

Claire Elizabeth Suver

Paul David Franklin

Mark Christopher Seigle

Nguyen Nguyen

Sandesh Doddameti

Rishabh Animesh

Jason Neil Yakimovich

Techniques described and suggested herein include distributed deletion request processing and verification. For example, incident to migration of original data from a first data store to a second data store, verifications and confirmations related to removing the original data from the first data store may be performed so as to ensure the integrity of the original data represented on the second data store prior to removing the actual original data on the first data store. In some embodiments, the verifications and confirmations performed in connection with a deletion request may be apportioned to multiple entities, each of which may not fully trust the others. As a result, in some embodiments, a given deletion request may only be fulfilled if all of the entities involved in the verification process individually provide authorization to execute the deletion request.

Techniques for distributing verification checks incident to data deletion

A system implements techniques for processing data across multiple layers. In some embodiments, a distributed data processing system tracks processing states via a monotonic process, whereby each state is irreversibly entered and cannot be entered without a predecessor state being verifiably completed. In some of such embodiments, a multi-layer data storage system tracks and processes data deletions across disparate layers such that a deletion state can be positively deduced from the deletion states of each of the disparate layers.

Techniques for verifiably processing data in distributed computing systems

Items in a sequence are associated with items in a set of items to be visited, such as data objects stored by a data storage system. A scheduling algorithm schedules visitation of the items in a manner that is stateless, so that items can be removed or added to the set without affecting previously scheduled visitations of the remaining items. The scheduling algorithm also spreads visitations among the items in a manner that prevents visitations from being focused on clusters of the items.

Stateless scheduling

Techniques described and suggested herein include various methods and systems for verifying integrity of redundancy coded data, such as erasure coded data shards. In some embodiments, a quantity of redundancy coded data elements, hereafter referred to as data shards (e.g., erasure coded data shards), sufficient to reconstruct the original data element from which the redundancy coded data elements are derived, is used to generate reconstructed data shards to be used for checking the validity of analogous data shards stored for the original data element.

Single quorum verification of erasure coded data

Techniques described and suggested herein include implementations of caches and scalers to handle data storage requests, and storage event status requests associated with data storage requests, in a scalable fashion. For example, a data storage system, such as a data storage system implemented by a computing resource service provider in connection with providing an archival storage service or other data storage service, may be implemented to maintain a consistent response time and backend capability for incoming data storage requests, which may be a component of ensuring a consistent customer experience for customers of an associated service, with little or no regard to peaky or high data storage request rates observed by the implementing data storage system.

Growing community of inventors

Seattle, WA, United States of America

Ryan Charles Schmitt