Going back to the issue:
http://markmail.org/search/?q=Multinode+cluster+setup+issues#query:Multinode%20cluster%20setup%20issues+page:1+mid:qi45trv4fdfwugwf+state:results
And having gone over the steps pointed to in the response....I looked at the
TaskTracker, and on startup I see the following error:
WARN org.apache.hadoop.mapred.TaskTracker: TaskTracker's
totalMemoryAllottedForTasks is -1. TaskMemoryManager is disabled.
A search for the error didn't help. But the slave box had 8GB of ram and my
Xmx is set to:
<property>
<name>mapred.child.java.opts</name>
<value>-Xmx256m</value>
</property>
in mapred-site.xml.
I've tried the following values: 2G, 1G, 512M, 256M
None work.
Also I only have 2 Mappers and 1 Reduce task
Hassaan Moin Khan
