Configuring Hadoop in PATROL Console

This topic provides instructions to configure and to delete a configured Hadoop environment.

• • To configure a Hadoop environment
  • To delete a configured Hadoop environment

To configure a Hadoop environment

1. From the PATROL Console work area, right-click Hadoop node (HADOOP_MAIN application class)
2. Choose the Register Hadoop Environment KM menu command to display the Register Hadoop environment dialog box.
3. Use the Register Hadoop environment dialog box, to add a new Hadoop environment. 
4. Enter the information, as follows:

   1. Environment name - Enter a logical unique name for the environment. A container will be created with the environment name. This environment will contain the Hadoop components. For example, if you add Hadoop NameNode details, you will see all the Hadoop NameNode components and data related to it. BMC recommends you to provide only alphanumeric characters in environment name. Valid values: a-z, A-Z, 0-9

      Invalid values: # ' . | ? \ " [ ] + = &

   2. Hadoop connection details, 
      1. Hadoop Vendor -  Select the monitored Hadoop vendor. The supported vendors are:
         • Apache Hadoop
         • IBM BigInsights
         • MapR
         • Cloudera
         • Horton
      2. Hadoop Host - Enter a unique, user-defined host name or an IP address.
      3. Hadoop Port - Enter the Hadoop port number. This port number depends on the component you want to monitor. If you want to monitor ResourceManager, enter the ResourceManager port number. Default port number is NameNode port, 50070.
      4. Hadoop Authentication - Select the authentication from the following:
         
         1. 
            
            1. None
            2. Basic
            3. Kerberos
      5. Hadoop User - Enter the username to connect to the Hadoop server. Depending on the Hadoop environment, the Hadoop User may not be required.
      6. Hadoop Password - Enter the password (if required) to connect to the Hadoop server.
   3. Kerberos authentication details

      1. krb5 file path - Enter the Kerberos file path.
         The krb5.conf file contains Kerberos configuration information. This includes information describing the default Kerberos realm, and the location of the Kerberos key distribution centers for known realms, and mappings of hostnames onto Kerberos realms. 

         Note

         The following are the steps required to monitorhadoopwithkerberos authentication:

         1. Select Kerberos under Hadoop Authentication section.
         2. The Kerberos username and password fields are required.
         3. Enter the full path of your hadoop krb5.conf file.
   4. HTTPS secure details
      
      1. Secure (HTTPS) - Select this checkbox if you want to use HTTPs protocol for connection. By default, HTTP protocol is used for a connection.
      2. Trust Store - If you select to use HTTPs protocol for connection, you must provide the location to the Java Keystore File (JKS) for a successful connection to the URL in a secure mode.
      3. Trust Password - Enter the password to the JKS file.
         
   5. Enter the Java collector settings,
      1. JAVA_HOME - Enter the directory path where you have installed Java version 1.7 or later. 
      2. JVM Arguments - Enter the JVM arguments for the Hadoop java collector.

1. 1. 1. Executing user - Enter the username that has permissions to run Java, only if the monitoring user does not have permissions.

1. 1. 1. Executing password - Enter the executing user password.
   2. Enter the Collection details,
      1. Availability -  Enables monitoring of the Hadoop monitored groups availability (if an instance is running or is down) of one of the following instances existing in the environment:
         1. 
            
            1. NameNode
            2. SecondaryNameNode
            3. ResourceManager
            4. DataNode
            5. NodeManager
            6. JobHistory
            7. JobTracker
            8. TaskTracker
            9. JournalNode
      2. Full - Enables the monitoring of all Hadoop monitored groups, like UGIMetrics, StartUpProgress, and other groups.

      3. Custom -  Manually select the Hadoop monitored groups to monitor. A table of the custom monitoring option is given below:

         Monitored group	Metrics
         Data Node	Select the list of metrics for monitoring: Metrics System Stats DataNode Activity FS Dataset State JVM Metrics RPC Activity DataNode Info UGI Metrics
         Job Tracker	Select the list of metrics for monitoring: RPC Activity RPC Detailed Activity Metrics System Stats Queue Metrics Default UGI Metrics JVM Metics Startup Progress JobTracker Metrics JobTracker Info
         Name Node	Select the list of metrics for monitoring: Metrics System Stats NameNode Activity RPC Detailed Activity NameNode Retry Cache FS Namesystem JVM Metrics NameNode Info UGI Metrics RPC Activity FS Namesystem State Startup Progress
         Node Manager	Select the list of metrics for monitoring: RPC Activity JVM Metrics NodeManager Metrics Shuffle Metrics RPC Detailed Activity UGI Metrics Metrics System Stats
         Resource Manager	Select the list of metrics for monitoring: RPC Detailed Activity JVM Metrics Metrics System Stats Cluster Metrics RPC Activity UGI Metrics Queue Metrics
         Secondary Name Node	Select the list of metrics for monitoring: UGI Metrics SecondaryNameNode Info Metrics System Stats JVM Metrics Startup Progress
         Task Tracker	Select the list of metrics for monitoring: Shuffle Server Metrics TaskTracker Metrics Metrics System Stats Startup Progress JVM Metrics TaskTracker Info RPC Activity UGI Metrics
         Note Due to PATROL Console limitations, if the collection details is set to Custom monitoring, you will not be able to select the JobHistory and JournalNodes groups for monitoring.
         JobHistory	Select the list of metrics for monitoring: Metrics System Stats RPC Activity JVM Metrics UGI Metrics
         JournalNode	Select the list of metrics for monitoring: Journal UGI Metrics JVM Metrics Metrics System Stats RPC Activity RPC Detailed Activity

2. Click OK to save the configuration.
3. Click Cancel to exit without saving.

To delete a configured Hadoop environment

1. From the PATROL Console work area, right-click Environment name that you want to delete.
2. Choose the Unregister Hadoop Environment KM menu command to delete the selected environment.
3. Click Ok to confirm. 

Related topics

Configuring-in-Central-Monitoring-Administration-console