Amazon Technologies, Inc.

Marco Losito

Jamie Plenderleith

Donagh Edward Hatton

Michael Achola

Marios Dimitriou

Edouard Stephane Gilles Marie Vincent

A multi-layered parallel aggregation can be performed on large-scale metric streams using layers of independent host server computers that perform partial aggregations on results of a previous layer and pass a result of the partial aggregation to a next layer of host server computers, until a single host server computer in a last layer can calculate a final output, which is a combination of the partial aggregations. Because every layer is aggregating input streams, the quantity of data exchange lowers as the layer number increases. The number of layers is chosen so as to ensure that data ingested by a last layer is sufficiently manageable that a single host server computer in the last layer can aggregate all of the partial aggregates into a final result.

Multi-layer network for metric aggregation

Technologies are provided for identifying a source of errors within a computer network that uses cut-through forwarding of network packets. For example, inbound network packets can be received by network devices configured to perform cut-through forwarding. Once the network packets are received, they are processed using cut-through forwarding. If any packet errors are detected, error data associated with the detected errors are stored by the network devices. The error data is transmitted to a monitoring service that uses the error data to create a model of a flow of the errors through the network and to identify a source of the errors. In at least some embodiments, a topology of the network is used by the monitoring service to create the model of the packet error flow.

Error source identification in cut-through networks

Techniques for detecting a predetermined waveform pattern in a signal received at a system are disclosed herein. A standard deviation and an average of the signal are calculated and used to determine a plurality of zones for the signal. A plurality of samples is obtained from the signal and one of the zones is associated with each of the samples. Transitions of the signal between the zones are identified based on the association of the zones with the samples. A predetermined waveform pattern may be detected in the signal based on the transitions between the zones. A notification may be generated indicating that the predetermined waveform pattern is detected based on a correspondence of the transitions of the signal matching a set of the predetermined transitions.

Growing community of inventors

Dublin, Ireland

Marco Losito