Abstract: A system for processing data is provided. During operation, the system obtains a current window of one or more intervals of timeseries data collected from a monitored system. Next, the system continuously performs a statistical hypothesis test that compares the one or more intervals of the time-series data with baseline values from historic time-series data associated with the monitored system. When the statistical hypothesis test indicates a deviation of the time-series data from the baseline values, the system outputs an alert of an anomaly represented by the deviation.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a set of components of a time-series performance metric associated with an anomaly in a performance of one or more monitored systems. For each component in the set of components, the system performs a statistical hypothesis test on the component to assess a deviation of the component from a baseline value of the component. When the statistical hypothesis test identifies a statistically significant deviation of the component from the baseline value, the system outputs an alert comprising a root cause of the anomaly that is represented by the statistically significant deviation of the component from the baseline value.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a component of a time-series performance metric associated with a server-side root cause of an anomaly in the time-series performance metric. Next, the system obtains a call graph representation of the component, wherein the call graph representation includes a parent node having a parent value of the component and a set of child nodes of the parent node, each child node having a corresponding child value of the component. The system then analyzes the call graph representation to identify one or more of the child nodes as sources of the anomaly. Finally, the system outputs an alert that identifies the sources of the anomaly.

Abstract: Techniques are provided for automatically accounting and apportioning of observed service execution times. In an embodiment, the computers determine a total amount of time to process a request. The computers determine which services are called in order to process the request. The called services include a first service and a second service. The computers determine a first amount of time that the first service spends in response to processing the request and a second amount of time that the second service spends in response to processing the request. The computers calculate a first percentage for the first service based on the first amount of time and the total amount of time and a second percentage for the second service based on the second amount of time and the total amount of time.

Abstract: The disclosed embodiments provide a system for detecting and managing inefficiency in external services. During operation, the system obtains a snapshot of a data stream transmitted over an external service from a computer system at a first time. Next, the system applies a compression technique to the snapshot to obtain a set of inefficiency metrics for a set of data elements in the snapshot. The system then outputs the set of inefficiency metrics with additional attributes of the data stream to improve identification of inefficiency in the data stream.

Abstract: The disclosed embodiments provide a system for detecting and managing inefficiency in local storage. During operation, the system obtains a first snapshot of data in local storage of a computer system, wherein the first snapshot comprises a first set of data elements in the local storage at a first time. Next, the system applies a compression technique to the first snapshot to obtain a first set of inefficiency metrics for the first set of data elements. The system then outputs the first set of inefficiency metrics with additional attributes of the data to improve management of inefficiency in the data.

Abstract: Techniques for automatically recommending a new data center, such as a point-of-presence (POP) center are provided. In one technique, multiple candidate locations for a new POP center are considered. An impact of adding a new POP to each candidate location is estimated and a score is generated. Each candidate location is ranked based on the score. In a related technique, an impact score for a candidate location is based on a prediction of whether and how much a new POP center at the candidate location would reduce response times of clients that would connect to the new POP center. The prediction may be based on a model that is generated based on response data generated by clients that are connecting to existing POP centers. The model may take into account multiple factors, such as geographic distance, network distance, type of browser executing on the clients, type of operating system, etc.

Abstract: Techniques are provided for analyzing client-side performance data. In one technique, first client performance data and second client performance data is received. The first client performance data was generated by a first client device and indicates a first plurality of processing items that includes a first processing item that is associated with first item performance data. The second client performance data was generated by a second client device and indicates a second plurality of processing items that includes a second processing item that is associated with second item performance data. In response to determining that the first processing item is associated with the second processing item, an aggregate data value is computed based on the first item performance data and the second item performance data.

Abstract: The disclosed embodiments provide a system for detecting and managing inefficiency in external services. During operation, the system obtains a snapshot of a data stream transmitted over an external service from a computer system at a first time. Next, the system applies a compression technique to the snapshot to obtain a set of inefficiency metrics for a set of data elements in the snapshot. The system then outputs the set of inefficiency metrics with additional attributes of the data stream to improve identification of inefficiency in the data stream.

Abstract: The disclosed embodiments provide a system for detecting and managing inefficiency in local storage. During operation, the system obtains a first snapshot of data in local storage of a computer system, wherein the first snapshot comprises a first set of data elements in the local storage at a first time. Next, the system applies a compression technique to the first snapshot to obtain a first set of inefficiency metrics for the first set of data elements. The system then outputs the first set of inefficiency metrics with additional attributes of the data to improve management of inefficiency in the data.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a set of components of a time-series performance metric associated with an anomaly in a performance of one or more monitored systems. For each component in the set of components, the system performs a statistical hypothesis test on the component to assess a deviation of the component from a baseline value of the component. When the statistical hypothesis test identifies a statistically significant deviation of the component from the baseline value, the system outputs an alert comprising a root cause of the anomaly that is represented by the statistically significant deviation of the component from the baseline value.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a component of a time-series performance metric associated with a server-side root cause of an anomaly in the time-series performance metric. Next, the system obtains a call graph representation of the component, wherein the call graph representation includes a parent node having a parent value of the component and a set of child nodes of the parent node, each child node having a corresponding child value of the component. The system then analyzes the call graph representation to identify one or more of the child nodes as sources of the anomaly. Finally, the system outputs an alert that identifies the sources of the anomaly.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a current window of one or more intervals of time-series data collected from a monitored system. Next, the system continuously performs a statistical hypothesis test that compares the one or more intervals of the time-series data with baseline values from historic time-series data associated with the monitored system. When the statistical hypothesis test indicates a deviation of the time-series data from the baseline values, the system outputs an alert of an anomaly represented by the deviation.

Abstract: A system, method, and apparatus are provided for assigning or allocating multiple content objects, within a content page (e.g., web page) or other content collection (e.g., a set of pages), to different content delivery networks for delivery in response to a content request. The objects are ranked by importance (e.g., importance in rendering or presenting the page), and the networks are ranked by performance (e.g., throughput). In order of importance, the objects are assigned to the best-performing network that is “available.” Some or all networks are initially available, and a given network becomes “unavailable” after it has been assigned its portion of the objects (e.g., based on content, number of objects, amount of data, percentage). If a total accumulated cost of delivering the objects exceeds a target before all objects have been allocated, the allocation process may terminate early and the remaining objects may be assigned to the least-expensive network.

Abstract: Techniques for reassigning client devices to different POP centers are provided. Client devices execute code that causes the client devices to request data from one or more POP centers. The client device record how much time it takes to receive the requested data. The download time is reported and aggregated to determine POP center-to-client device assignments. If it is determined that a client device should be reassigned to a particular POP center, then a DNS provider is configured to update the appropriate hostname-to-IP address mapping to an IP address of the particular POP center. A change in performance based on a POP reassignment may then be determined.

Abstract: Techniques for automatically recommending a new data center, such as a point-of-presence (POP) center are provided. In one technique, multiple candidate locations for a new POP center are considered. An impact of adding a new POP to each candidate location is estimated and a score is generated. Each candidate location is ranked based on the score. In a related technique, an impact score for a candidate location is based on a prediction of whether and how much a new POP center at the candidate location would reduce response times of clients that would connect to the new POP center. The prediction may be based on a model that is generated based on response data generated by clients that are connecting to existing POP centers. The model may take into account multiple factors, such as geographic distance, network distance, type of browser executing on the clients, type of operating system, etc.

Abstract: Techniques for evaluating the performance of a service or machine after a change that pertains to the service or machine are described. For example, an indication of a change that pertains to a service or machine is received. In response to the receiving of the indication of the change, using at least one computer processor, a performance of the service or machine after the change is evaluated. The evaluation may be based on a particular rule for evaluating the performance of the service or machine after the change. An evaluation result is generated based on the evaluating of the performance of the service or machine after the change. The evaluation result indicates the quality of the performance of the service or machine after the change.