MapReduce programming Series 7 MapReduce program log view, mapreduce log

MapReduce programming Series 7 MapReduce program log view, mapreduce log

First, to print logs without using log4j, you can directly use System. out. println. The log information output to stdout can be found at the jobtracker site.

Second, if you use System. out. println to print the log when the main function is started, you can see it directly on the console.

Second, the jobtracker site is very important.

Http: // your_name_node: 50030/jobtracker. jsp

Note: here we can see that Map 100% is not necessarily correct. Sometimes it is stuck in the Map stage and it is not completed yet. At this time, Map 100% is displayed, so we need to click it layer by layer until we can see the log.

In addition, the map/reduce slots can be seen in the cluster summary table to facilitate understanding of cluster computing resources. You can write a script to collect slots information on a regular basis to analyze the cluster peak and idle time periods.

How to record hadoop mapreduce running logs

In the logs directory of the master and slaves machines.

In the cloud computing environment, how can we use the mapreduce programming model to program and then put the program on the cluster for running?

First, write the mapreduce code based on your specific application, then package the mapreduce code and run it on the hadoop cluster that has been built.