Akamai Technologies, Inc.

Aniruddha Bohra

Mithila Nagendra

Vadim Grinshpun

Hari Raghunathan

Robert Bird

Vinay Kanitkar

Michael G Merideth

Umberto Boscolo Bragadin

James Lee

Solomon Lifshits

Florin Sultan

This patent document describes failure recovery technologies for the processing of streaming data, also referred to as pipelined data. The technologies described herein have particular applicability in distributed computing systems that are required to process streams of data and provide at-most-once and/or exactly-once service levels. In a preferred embodiment, a system comprises many nodes configured in a network topology, such as a hierarchical tree structure. Data is generated at leaf nodes. Intermediate nodes process the streaming data in a pipelined fashion, sending towards the root aggregated or otherwise combined data from the source data streams towards. To reduce overhead and provide locally handled failure recovery, system nodes transfer data using a protocol that controls which node owns the data for purposes of failure recovery as it moves through the network.

Systems and methods for failure recovery in at-most-once and exactly-once streaming data processing

Described in this document, among other things, is an overload protection system that can protect data sinks from overload by controlling the volume of data sent to those data sinks in a fine-grained manner. The protection system preferably sits in between edge servers, or other producers of data, and data sinks that will receive some or all of the data. Preferably, each data sink owner defines a policy to control how and when overload protection will be applied. Each policy can include definitions of how to monitor the stream of data for overload and specify one or more conditions upon which throttling actions are necessary. In embodiments, a policy can contain a multi-part specification to identify the class(es) of traffic to monitor to see if the conditions have been triggered.

Overload protection for data sinks in a distributed computing system

Edge server compute capacity demand in an overlay network is predicted and used to pre-position compute capacity in advance of application-specific demands. Preferably, machine learning is used to proactively predict anticipated compute capacity needs for an edge server region (e.g., a set of co-located edge servers). In advance, compute capacity (application instances) are made available in-region, and data associated with an application instance is migrated to be close to the instance. The approach facilitates compute-at-the-edge services, which require data (state) to be close to a pre-positioned latency-sensitive application instance. Overlay network mapping (globally) may be used for more long-term positioning, with short-duration scheduling then being done in-region as needed. Compute instances and associated state are migrated intelligently based on predicted (e.g., machine-learned) demand, and with full data consistency enforced.

Intelligently pre-positioning and migrating compute capacity in an overlay network, with compute handoff and data consistency

Described in this document, among other things, is an overload protection system that can protect data sinks from overload by controlling the volume of data sent to those data sinks in a fine-grained manner. The protection system preferably sits in between edge servers, or other producers of data, and data sinks that will receive some or all of the data. Preferably, each data sink owner defines a policy to control how and when overload protection will be applied. Each policy can include definitions of how to monitor the stream of data for overload and specify one or more conditions upon which traffic shaping actions are necessary. In embodiments, a policy can contain a multi-part specification to identify the class(es) of traffic to monitor to see if the conditions have been triggered.

Growing community of inventors

Philadelphia, PA, United States of America

Aniruddha Bohra