Abstract: A web site provides services to uses over the Internet. End-user service-level objectives (SLOs) such as availability and performance are measured and reported to a SLO agent. User requests pass through several tiers at the web site, such as a firewall tier, a web-server tier, an application-server tier, and a database-server tier. Each tier has several redundant service components that can process requests for that tier. Local agents, operating with any local resource managers, monitor running service components and report to a service agent. Node monitors also monitor network-node status and report to the service agent. When a node or service component fails, the service agent attempts to restart it using the local agent, or replicates the service component to other nodes. When the SLO agent determines that a SLO is not being met, it instructs the service agent to replicate more of the constraining service components or increase resources.

Abstract: A system and method for monitoring a set of performance metrics is described. The methods include metric correlation and grouping methods that analyze a group of temporally related metrics, and correlate pairs of the metrics in the group. In one embodiment, rank correlation techniques are used to perform this correlation. Methods are also described for grouping metrics using a dynamic correlation pair graph that preserves all of the correlated relationships. This correlation pair graph may be used to determine which metrics are associated with a particular key metric, which may provide information on the cause of an alarm or other event involving the key metric. Another embodiment includes apparatus for correlating, grouping, and identifying the metrics.

Abstract: A system and method for monitoring a set of performance metrics is described. The methods include metric correlation and grouping methods that analyze a group of temporally related metrics, and correlate pairs of the metrics in the group. In one embodiment, rank correlation techniques are used to perform this correlation. Methods are also described for grouping metrics using a dynamic correlation pair graph that preserves all of the correlated relationships. This correlation pair graph may be used to determine which metrics are associated with a particular key metric, which may provide information on the cause of an alarm or other event involving the key metric. Another embodiment includes apparatus for correlating, grouping, and identifying the metrics.

Abstract: A system and methods for displaying data distribution information for time-series data is described. The methods include computing a condensed quantile function that may be used to generate approximate histograms for the time-series data, while decreasing the data storage requirements for generating a series of histograms for time-series data. The methods further include displaying the data distribution information using stack-bar histograms, many of which may be shown in a single display to permit a user to discern trends in the data distribution information. Methods for merging condensed quantile function tables are also described.

Abstract: A system and method for dynamically generating alarm thresholds for performance metrics, and for applying those thresholds to generate alarms is described. Statistical methods are used to generate one or more thresholds for metrics that may not fit a Gaussian or normal distribution, or that may exhibit cyclic behavior or persistent shifts in the values of the metrics. The statistical methods used to generate the thresholds may include statistical process control (SPC) methods, normalization methods, and heuristics.