Logging in Hadoop Stream jobs

[Mayuran Yogarajah]

How do people handle logging in a Hadoop stream job?

I'm currently looking at using syslog for this but would like to know of 
other ways
people are doing this currently.
thanks