Airflow

Airflow允许您使用有向无环图（DAG）将任务组织成工作流。Airflow调度器会按照指定的依赖关系在工作节点上执行您的任务。丰富的命令行实用程序使您可以轻松地对DAG进行复杂操作。用户界面使您可以轻松地可视化正在生产中运行的管道，监控进度并在需要时解决问题。

Airflow 可以将工作流定义为有向无环图 (DAG)，并按照依赖关系在多个 worker 上调度和执行任务。其强大的命令行工具和丰富的用户界面，为用户提供了便捷的 DAG 管理、流程可视化、进度监控和问题排查功能。

本指南介绍了如何使用最新版 Airflow 设置中国时区，并附有相关中文资料和详细的安装方案。Airflow 是一个用于编排、调度和监控工作流的平台，它将工作流编排为 DAG，由调度器执行任务。同时，它还提供了命令行工具、用户界面、监控和告警系统。

The FLO MATLAB code is a tool designed for tracking airflow with detectable tracers. The tracer used in our lab is a laser-illuminated bubble. The code processes videos captured of airflow with the tracers, and traces the trajectory of the bubbles. This allows you to determine various parameters such as the velocity of the airflow at different locations and the curvature of the flow. To run the code, simply add the FLO directory to your MATLAB path. Then, run the script BT_Driver located in the FLO/src/processingFiles/ folder. To modify the video or change the parameters, adjust the values in the script. After running BT_Driver, the processed data will be stored in the workspace variable S. To view the tracks found by the script, use the command plotTracks(S, frameNumber, 1);. Here, S is the variable created by BT_Driver. To see the position of each bubble in the video, run the following: cen = allCentriod(S); plot(cen(:,1), cen(:,2), '.');. For a density map of the bubbles detected in the video, use: imagesc(bin2d(cen, binSize, [vidDimX, vidDimY])); colorbar();.

该项目利用Sqoop、Hive、Spark、MySQL、AirFlow和Grafana技术，构建了一个工业大数据的离线数据仓库。