Panappticon: Event-based tracing to measure mobile application and platform performance

Abstract

Improving and optimizing user-perceived smartphone performance requires understanding device, system, and application behavior for real-world workloads. However, measuring such performance is challenging due to the multi-threaded, asynchronous programming paradigms used in modern applications and the multiple layers of hardware and software used to respond to user input events. We address this challenge with Panappticon, a lightweight, system-wide, fine-grained event tracing system for Android that automatically identifies critical execution paths in user transactions. Panappticon monitors the application, system, and kernel software layers and can identify performance problems stemming from application design flaws, underpowered hardware, and harmful interactions between apparently unrelated applications. We carried out a 14-user, one-month study of an Android smartphone system instrumented with Panappticon, which revealed a number of specific problems and areas for improvement that may be of interest to system designers, application developers, and device manufactures.

Diagnosing Energy Efficiency and Performance for Mobile Internetware Applications: Challenges and Opportunities

“…AppInsight [11] can help instrument the binaries of smartphone applications to identify long latency execution paths. Panappticon [15] further supports revealing performance issues arising from inefficient platform code or problematic interactions among applications. The most recent work SunCat [5] logs the events in a test run and summarizes event repetition patterns to help developers understand and predict performance problems.…”

“…Third, engineering experience may help make judgments. For instance, many developers consider an application suffering from performance bugs if a user event cannot be handled within 200 milliseconds [2,15]. Although these strategies are widely-used, they either require non-trivial manual effort or are not generally defined.…”

“…Recent studies reported that these bugs had a severe impact on user experience and caused significant user frustration [7,15]. Despite the pervasiveness of such bugs, it is unfortunate that both research communities and industries lack a good understanding of these bugs.…”

QoE Doctor

“…Thus, it is the main focus of this paper. Unlike previous work [39,15,47], our measurement approach (described in §4) directly calculates the latency from user's perspective -the UI layer, without requiring application source code or any OS/application logic instrumentation, which enables us to study this QoE metric broadly on any popular mobile apps of interest.…”

User interaction-based profiling system for Android application tuning

“…In [16,17], operation path is analyzed from the point when the user provides an input. Its corresponding output is shown on the display, and the longest path is shown to reduce response time.…”