I just noticed that "mapreduce.job.split.metainfo.maxsize" is given in
mapred-default.xml while the code in
public static final String JT_MAX_JOB_SPLIT_METAINFO_SIZE =
I'm looking at hadoop-mapreduce trunk as pulled from SVN earlier today.
I'd be happy to apply this myself if I can figure out the procedure: Create
a JIRA and submit a patch? Do I need to be granted access to create a JIRA?