Cisco Technology, Inc.

Nishant Agarwal

Rajesh Raman

Joseph Ari Ross

Techniques used by an observability system for facilitating near real-time analytics using structures representing or summarizing data distributions, such as histograms. An observability system receives a set of datapoints in which each received datapoint includes data for a structure representing a data distribution. For example, each of the datapoints may include data for a histogram including at least two dimensions. A processing pipeline is disclosed for processing the received data and making the results of the processing available to a user in near real-time. The processing includes identifying, from the datapoints received by the observability system, a subset of datapoints that fall within a time window, generating aggregate data based upon data in the identified subset of datapoints, computing one or more statistical measures based upon the aggregate data, and enabling querying of the subset of datapoints, the aggregate data, and the statistical measures computed for the first minute.

Growing community of inventors

Santa Clara, CA, United States of America

Nishant Agarwal