Search

Distributed systems are complex to develop and administer, and performance problem diagnosis is particularly challenging. When performance degrades, the problem might be in any of the system's many components or could be a result of poor interactions among them. Recent research efforts have created tools that automatically localize the problem to a small number of potential culprits, but research is needed to understand what visualization techniques work best for helping distributed systems developers understand and explore their results. This paper compares the relative merits of three well-known visualization approaches (side-by-side, diff, and animation) in the context of presenting the results of one proven automated localization technique called request-flow comparison. Via a 26-person user study, which included real distributed systems developers, we identify the unique benefits that each approach provides for different problem types and usage modes.

Metrics like disk activity and network traffic are widespread sources of diagnosis and monitoring information in datacenters and networks. However, as the scale of these systems increases, examining the raw data yields diminishing insight. We present RainMon, a novel end-to-end approach for mining timeseries monitoring data designed to handle its size and unique characteristics. Our system is able to (a) mine large, bursty, real-world monitoring data, (b) find significant trends and anomalies in the data, (c) compress the raw data effectively, and (d) estimate trends to make forecasts. Furthermore, RainMon integrates the full analysis process from data storage to the user interface to provide accessible long-term diagnosis. We apply RainMon to three real-world datasets from production systems and show its utility in discovering anomalous machines and time periods.

Mobile phone services based on the location of a user have increased in popularity and importance, particularly with the proliferation of feature-rich smartphones. One major obstacle to the widespread use of location-based services is the limited battery life of these mobile devices and the high power costs of many existing approaches.We demonstrate the effectiveness of a localization strategy that performs full localization only when it detects a user has finished moving. We characterize the power use of a smartphone, then verify our strategy using models of long-term walk behavior, recorded data, and device implementation. For the same sample period, our movement-informed strategy reduces power consumption compared to existing approaches by more than 80% with an impact on accuracy of less than 5%. This difference can help achieve the goal of near-continuous localization on mobile devices.