Suppose you've created HDFS in ~/. Then the log files are in ~/hadoop_data/hadoop_local-YOURUSER/userlogs/job_yyyymmdd_xxxx/attempt_yyyymmdd_xxxx_*/log-file where:
- yyyymmdd is the date you started Hadoop - xxxx is the job number - attempt_* will be different for different workers - log-file can be stderr, stdout, or syslog Typically you want to look at syslog for anything printed using log4j. Metrics, if enabled, will show up in stderr. If you're running distributed, each machine will log things to their local disk. You can also get to the logs by going to (assuming your Hadoop files are in ~/) ~/hadoop-x.y.z/logs/userlogs: the folders in here contain symlinks to the directories above. Young On Tue, May 12, 2015 at 11:26 AM, Cheng Wang <[email protected]> wrote: > Hello, > > I am sure if it is good to ask here.. I am new to Giraph. > I found that in the example codes provided by Giraph package, there are > many outputs are stored into log4j files... e.g. > SimplePageRankComputation.java. > > I am wondering where are the log files stored in the log4j? > > > Thanks > Cheng >
