Database Reference
In-Depth Information
You can drill down on the data nodes, access their file system, and go all the way to job configurations used
during job submission, as shown in Figure 6-14 .
Figure 6-14. The job configurations
The Name Node Status Portal is a part of the Apache Hadoop project, making it familiar to existing Hadoop users.
The main advantage of the portal is that it lets you browse through the file system as if it is a local file system. That's
an advantage because there is no way to access the file system through standard tools like Windows Explorer, as the
entire storage mechanism is abstracted in WASB.
The TaskTracker Portal
Apart from the Name Node and MapReduce status portals, there is also a TaskTracker web interface that is available
only in the data nodes or task nodes of your cluster. This portal listens on port 50060, and the complete URL to launch
it is http://<DataNode_IP_Address>:50060/tasktracker.jsp . Although there is a single TaskTracker per slave node,
each TaskTracker can spawn multiple JVMs to handle many map or reduce tasks in parallel.
Note
The TaskTracker service runs on the data nodes, so there is no shortcut created for that portal in the name node.
You need to log on remotely to any of your data nodes to launch the TaskTracker portal. Remember, the remote
logon session needs to be initiated from the name node Remote Desktop session itself. It will not work if you try to
connect remotely to your data node from your client workstation. This Java-based web portal displays the status of the
completed tasks along with their status, as shown in Figure 6-15 .
 
 
Search WWH ::




Custom Search