Systems, methods, and other embodiments associated with autonomous cloud-node scoping for big-data machine learning use cases are described. In some example embodiments, an automated scoping tool, method, and system are presented that, for each of multiple combinations of parameter values, (i) set a combination of parameter values describing a usage scenario, (ii) execute a machine learning application according to the combination of parameter values on a target cloud environment, and (iii) measure the computational cost for the execution of the machine learning application. A recommendation regarding configuration of central processing unit(s), graphics processing unit(s), and memory for the target cloud environment to execute the machine learning application is generated based on the measured computational costs.

A test and measurement instrument includes an acquisition memory and a processor structured to store a stream of sampled incoming data samples in the acquisition memory. As the memory fills, the instrument automatically decimates either the data samples already stored in the acquisition memory, the incoming data samples, or both. The instrument may also store two copies of the incoming data samples, one at an increased decimation rate. The two copies are tied together with a timestamp or using other methods. The more highly decimated copy may be used to produce a video output of the stored data samples, saving the instrument from generating the video output from the larger sized sample.

Technology is described for receiving time series data to be transmitted to a server. A network connectivity problem may be determined to exist for a computer network with the server which prevents the time series data from being transmitted to the server. A downsampling function may be applied to the time series data to produce reduced granularity data points that represent an approximation of the time series data to be transmitted to the server after the network connectivity problem has occurred with the server. The reduced granularity data points may be transmitted to the server.

A device receives network constraints associated with a network connected to a vehicle device, data collection constraints, and vehicle device constraints. The device determines a first sampling rate, a first time period, a second sampling rate, and a second time period for collecting vehicle data based on the network constraints, the data collection constraints, and the vehicle device constraints, wherein the first sampling rate is different than the second sampling rate. The device receives first vehicle data, provided at the first sampling rate and for the first time period, and second vehicle data, provided at the second sampling rate and for the second time period, and processes the first and second vehicle data, with a machine learning model, to identify a driving behavior or a vehicle issue. The device performs one or more actions based on the driving behavior or the vehicle issue.

An information processing device capable of reducing an amount of data to be monitored in an onboard system is provided. The information processing device obtains a first log including, per unit time, some communication data flowing through the onboard system. The information processing device determines whether an abnormality is included in the communication data, using the first log. In a case where the abnormality is included in the communication data, the information processing device outputs first detection results to the onboard system. The first detection results cause transmission of a second log from the onboard system. The second log includes, per the unit time, more of the communication data than the first log.

Systems and methods are disclosed for drop detection and protection with respect to packet monitoring in virtual processing environments. Tap agents monitor and capture packets from the network traffic associated with network applications running within these virtual processing environments. Sequence numbers are added in packet encapsulation before tap packets are forwarded to tool agents. The tool agents then use the sequence numbers to detect packet drops within the tap packets. After drop detection, the tool agents send drop detection messages to an agent controller, and the agent controller generates and sends reconfiguration messages to the tap agents based upon the drop detection messages. The tool agents can also send drop detection messages directly to the tap agents. The tap agents adjust their operations based upon the reconfiguration messages and/or the drop detection messages to reduce packet drops within subsequent tap packets communications.

Processes and systems described herein are directed to determining efficient sampling rates for metrics generated by various different metric sources of a distributed computing system. In one aspect, processes and systems retrieve the metrics from metric data storage and determine non-constant metrics of the metrics generated by the various metric sources. Processes and systems separately determine an efficient sampling rate for each non-constant metric by constructing a plurality of corresponding reduced metrics, each reduced metric comprising a different subsequence of the corresponding metric. Information loss is computed for each reduced metric. An efficient sampling rate is determined for each metric based on the information losses created by constructing the reduced metrics. The efficient sampling rates are applied to corresponding streams of run-time metric values and may also be used to resample the corresponding metric already stored in metric data storage, reducing storage space for the metrics.

A system and method for determining the traceability of network request traffic over a communications network for reducing strain in traffic processing resources, which includes: provisioning a direct interconnect on the communications network between the server and a predefined source, the direct interconnect providing a private service interface, a defined pairings data of the predefined source with the direct interconnect stored as a network traffic almanac; provisioning a public service interface on the communications network; receiving a request traffic having an address of the predefined source via the public service interface; consulting the defined pairing data with the address to determine the request traffic matches the predefined source; and de-prioritizing the processing of the request traffic based on the request traffic being received on the public service interface rather than on the direct interconnect by dynamically applying a prioritize criterion to the second request traffic before generating a response traffic.

A mechanism for adaptively performing in-band network telemetry (INT) by a network controller is disclosed. The mechanism includes receiving one or more congestion indicators from a collector. An adjusted sampling rate is generated. The adjusted sampling rate is a specified rate of insertion of instruction headers for INT and is generated based on the congestion indicators. The adjusted sampling rate is transmitted to a head node, which is configured to perform INT via instruction header insertion into user packets.

A method and system of accelerating monitoring of network traffic. The method may include receiving, at a network chip of a network device, a network traffic data unit; capturing, by the network chip, the network traffic data unit based on a traffic sampling rate; adding, by the network chip, a sampling header to the network traffic data unit to obtain a sampled network traffic data unit; sending the sampled network traffic data unit from the network chip to a sampling engine; receiving, from the sampling engine, a flow datagram that includes a network traffic data unit portion and a flow datagram header; generating a flow network data traffic unit that includes the flow datagram; and transmitting the flow network data traffic unit towards a collector.