NameNode Stats

Sample Image

Description

The NameNode holds the block information for the Hadoop File System. The NameNode provides counters for each operation type it performs. This information is useful to profile what the NameNode is doing and understanding how the framework utilizes this component.

Propaganda

Most of the work spent with Hadoop Cacti JTG has been spent grouping logically related counters into single graphs. Rather then looking at four separate graphs (that may scaled differently) users can look at one graph.

Sample Output

           sh /var/www/html/cacti-jtg/scripts/NameNodeStatistics.sh \
           service:jmx:rmi:///jndi/rmi://10.0.1.2:10001/jmxrmi controlRole password \
           BlockReportAverageTime:0 BlockReportMaxTime:5 BlockReportMinTime:0 \
           BlockReportNum:0 JournalTransactionAverageTime:0 JournalTransactionNum:0 \
           JournalTransactionMaxTime:1 JournalTransactionMinTime:0 \
           JournalSyncAverageTime:0 JournalSyncMaxTime:6 JournalSyncMinTime:0 \
           JournalSyncNum:0 SafemodeTime:39830 FSImageLoadTime:175 \
           NumFilesCreated:0 NumFilesListed:0 NumGetBlockLocations:0 NumFilesRenamed:0
          

Code Specific

Unlike some Hadoop variables that are based on a sample period these values when extracted from JMX represent the current state.

Continue to Datanode Capacity Graph