org.apache.hadoop.contrib.failmon
Class Executor

java.lang.Object
  extended by org.apache.hadoop.contrib.failmon.Executor
All Implemented Interfaces:
Runnable

public class Executor
extends Object
implements Runnable

This class executes monitoring jobs on all nodes of the cluster, on which we intend to gather failure metrics. It is basically a thread that sleeps and periodically wakes up to execute monitoring jobs and ship all gathered data to a "safe" location, which in most cases will be the HDFS filesystem of the monitored cluster.


Field Summary
static int DEFAULT_LOG_INTERVAL
           
static int DEFAULT_POLL_INTERVAL
           
static int instances
           
static int MIN_INTERVAL
           
 
Constructor Summary
Executor(Configuration conf)
          Create an instance of the class and read the configuration file to determine the set of jobs that will be run and the maximum interval for which the thread can sleep before it wakes up to execute a monitoring job on the node.
 
Method Summary
 void cleanup()
           
 void run()
           
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

DEFAULT_LOG_INTERVAL

public static final int DEFAULT_LOG_INTERVAL
See Also:
Constant Field Values

DEFAULT_POLL_INTERVAL

public static final int DEFAULT_POLL_INTERVAL
See Also:
Constant Field Values

MIN_INTERVAL

public static int MIN_INTERVAL

instances

public static int instances
Constructor Detail

Executor

public Executor(Configuration conf)
Create an instance of the class and read the configuration file to determine the set of jobs that will be run and the maximum interval for which the thread can sleep before it wakes up to execute a monitoring job on the node.

Method Detail

run

public void run()
Specified by:
run in interface Runnable

cleanup

public void cleanup()


Copyright © 2009 The Apache Software Foundation