Monitoring Data - Big Data Made Easy: A Working Guide to the Complete Hadoop Toolset - page 206

The next-to-last icon, which looks like an “H,” is for the HBase browser. Using this browser, you can examine
HBase tables, determine their structure, and manipulate their data. In addition, you can add rows or execute a bulk
upload of data (Figure 7-14 ).
Figure 7-14. The Hue HBase user interface
The right-most icon (the question mark) is the Hue help option, which is comprehensive. It has a series of sub-
help sections that include helpful topics on all of the Hadoop tools that have been integrated into Hue. If using Hue or
its components becomes unclear at any point, the Hue help pages are a good place to find extra information.
That completes a very quick tour of the Hue browser. It packs a lot of Hadoop functionality into a single user
interface, thereby enabling you to visually create and monitor Hadoop-based jobs and workflows from one interface.
What it doesn't offer, however, is any low-level monitoring of system-level resources. For that, you need Ganglia and
Nagios. The next sections present these monitoring tools, which can be used with a pre-existing Hadoop cluster.
Ganglia
An open-source monitoring system released under a BSD license, Ganglia is designed for monitoring on distributed
high-performance systems. With Ganglia installed you can monitor a Hadoop-based cluster. The official Ganglia
website is ganglia.sourceforge.net . Ganglia has been integrated into the Ambari Hadoop cluster manager, which
you will examine in Chapter 8.
While Hue provided a single, web-based user interface for accessing Hadoop-based components, Ganglia offers
true monitoring functionality. For example, it is possible to set up graph-based dashboards in Ganglia that show the
state of Hadoop cluster resources. At a glance, it is possible to determine if there is a cluster problem at the present
time or there has been one during the lifetime of each graph. Also, it is possible to add graphs for multiple types of
resources on a single dashboard. (Ganglia will again be discussed in regard to the Ambari Cluster manager, in the next
chapter.)
Here, I first tell how to source and install Ganglia. I also discuss some of the errors that might be encountered.
Then, I discuss its user interface.