Abstract: In some examples, time-series datasets received from a system may be temporally aligned. In some examples, one of the time-series datasets may be deduplicated. In some examples, whether an anomaly has occurred in the system may be determined based on a non-deduplicated time-series dataset of the time-series datasets.

Abstract: Event-time pairs are received for a current time slot. Each event-time pair denotes the occurrence of an event at a system by an event type as well as an occurrence time. For each different event type, a property value for the time slot is computed for each different property of a number of different properties, from the event-time pairs having the different event type. For each different property, a time-decaying histogram of identified property values of the different property is updated using the property value computed for the different property for the current time slot. An anomaly score for each identified property value within the time-decaying histogram of each different property is computed to detect occurrence of an anomaly within the system.

Abstract: An example process for aligning time-series datasets includes receiving a first time-series dataset and a second time-series dataset. The first time-series dataset can include a first set of values associated with respective time stamps and the second time-series dataset can include a second set of values associated with respective time stamps. The process also includes determining degrees of variance of the first and second sets of values, and comparing each degree of variance with a threshold. The process also includes selecting among multiple time alignment processes based on the comparisons, and processing the time-series datasets according to the selected process to thereby generate an aligned time-series dataset.

Abstract: In some examples, host IDs associated with the respective source component and a result of a partial calculation of an aggregate metric score may be received from each of a plurality of source components associated with a host of an information technology (IT) system. The partial calculation based on individual metric scores may be associated with the respective source component. The aggregate metric score may be calculated using the partial calculations and the host IDs, the aggregate metric score associated with metric measurements of the source components.

Abstract: In some examples, time-series datasets received from a system may be temporally aligned. In some examples, one of the time-series datasets may be deduplicated. In some examples, whether an anomaly has occurred in the system may be determined based on a non-deduplicated time-series dataset of the time-series datasets.

Abstract: Examples relate to calculating normalize metrics. The examples disclosed herein calculate respective normalized first metric values for each of a plurality of first metric values that are on a time scale and respective normalized second metric values for each of the plurality of raw second metric values that are on the time scale, where the plurality of first metric values are associated with a first metric, and the plurality of second metric values are associated with a second metric. An extremum of the normalized first metric value and the normalized second metric value at each time of the time scale is averaged to calculate a plurality of extremum baseline values. Examples herein calculate a plurality of sleeve values of the plurality of extremum baseline values based on a standard deviation of the plurality of extremum baseline values.

Abstract: Event-time pairs are received for a current time slot. Each event-time pair denotes the occurrence of an event at a system by an event type as well as an occurrence time. For each different event type, a property value for the time slot is computed for each different property of a number of different properties, from the event-time pairs having the different event type. For each different property, a time-decaying histogram of identified property values of the different property is updated using the property value computed for the different property for the current time slot. An anomaly score for each identified property value within the time-decaying histogram of each different property is computed to detect occurrence of an anomaly within the system.

Abstract: Examples relate to calculating normalize metrics. The examples disclosed herein calculate respective normalized first metric values for each of a plurality of first metric values that are on a time scale and respective normalized second metric values for each of the plurality of raw second metric values that are on the time scale, where the plurality of first metric values are associated with a first metric, and the plurality of second metric values are associated with a second metric. An extremum of the normalized first metric value and the normalized second metric value at each time of the time scale is averaged to calculate a plurality of extremum baseline values. Examples herein calculate a plurality of sleeve values of the plurality of extremum baseline values based on a standard deviation of the plurality of extremum baseline values.

Abstract: In one implementation, a system for resource consuming tasks scheduling includes a metrics engine to identify a plurality of metrics based on a task to be scheduled, a utilization engine to generate a compiled representation of combined resource values over a period of time, wherein the combined resource values are based on a combination of the plurality of metrics;, an identification engine to identify common lows of resource utilization for the plurality of metrics based on the generated representation, and a schedule engine to schedule the task for a time that corresponds to the identified common lows of resource utilization.

Abstract: An example process for aligning time-series datasets includes receiving a first time-series dataset and a second time-series dataset. The first time-series dataset can include a first set of values associated with respective time stamps and the second time-series dataset can include a second set of values associated with respective time stamps. The process also includes determining degrees of variance of the first and second sets of values, and comparing each degree of variance with a threshold. The process also includes selecting among multiple time alignment processes based on the comparisons, and processing the time-series datasets according to the selected process to thereby generate an aligned time-series dataset.