End-to-end Tracing in HDFS
نویسنده
چکیده
Debugging performance problems in distributed systems is difficult. Thus many debugging tools are being developed to aid diagnosis. Many require information from end-to-end tracing in order to perform their analysis. This paper describes the development of an end-to-end tracing framework for the Hadoop Distributed File System. The approach to instrumentation in this implementation differs from previous ones as it focuses on detailed low-level instrumentation. Such instrumentation encounters the problems of large request flow graphs and a large number of different kinds of graphs which impede the effectiveness of the diagnosis tools that use them. The paper describes how to instrument at a high granularity and explain techniques to handle the resulting challenges. The current implementation is evaluated in terms of performance, scalability, the data the instrumentation generates, and its ability to be used to solve performance problems.
منابع مشابه
Chukwa: A System for Reliable Large-Scale Log Collection
Large Internet services companies like Google, Yahoo, and Facebook use the MapReduce programming model to process log data. MapReduce is designed to work on data stored in a distributed filesystem like Hadoop’s HDFS. As a result, a number of log collection systems have been built to copy data into HDFS. These systems often lack a unified approach to failure handling, with errors being handled s...
متن کاملDiamond-forming Fluids and Kimberlites: The Trace Element Perspective
Major element studies of fluid inclusions in fibrous and cloudy diamonds reveal that most inclusions trapped high density fluids (HDFs) with a wide range of composition. The compositions fall along two arrays between three end-members: carbonatitic to hydroussilicic and carbonatitic to hydrous-saline. Recent studies of diamonds from Siberia and Kankan (Navon et al., 2008; Weiss et al., 2008) fu...
متن کاملEnd-to-end tracing considered essential
Concurrency and complexity are major obstacles to understanding application performance in high-performance systems. We advocate end-to-end event tracing as the correct way to expose performance information for both human and automated analysis. We describe its advantages over traditional performance counter data, and illustrate its uses in performance visualization, prediction for capacity pla...
متن کاملSo, youwant to trace your distributed system? Key design insights from years of practical experience
End-to-end tracing captures the workow of causally-related activity (e.g., work done to process a request) within and among the components of a distributed system. As distributed systems grow in scale and complexity, such tracing is becoming a critical tool for management tasks like diagnosis and resource accounting. Drawing upon our experiences building and using end-to-end tracing infrastruc...
متن کاملGeneral Baggage Model for End-to-End Tracing and Its Application on Critical Path Analysis
Many monitoring and diagnosis systems have been proposed based on causal tracing for end-to-end executions in distributed systems as more and more modern computer applications become distributed systems. These applications are based on metadata propagation along the request’s executions, which has been closely tied to the application logic, the metadata formats, and the developer APIs. Such kin...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2011