This is documentation for MapR Version 5.0. You can also refer to MapR documentation for the latest release.

Skip to end of metadata
Go to start of metadata

MapR provides several interfaces for working with a cluster from a client computer:

  • MapR Control System: Access the MapR Control System (MCS) to manage the cluster, including nodes, volumes, users, and alarms.
  • Direct Access NFS™: Configure Direct Access NFS to mount the MapR-FS to a local directory.
  • MapR client: Install the MapR client to run hadoop commands, jobs, and applications from a client machine.
  • MapR POSIX client: Enable app servers, web servers, and other client nodes and apps to read and write directly and securely to a MapR cluster.

MapR Control System

The MapR Control System allows you to control the cluster through a comprehensive graphical user interface.

Browser Compatibility

The MapR Control System is web-based, and works with the following browsers:

  • Chrome
  • Safari 
    • Version 5.1 and below with unsigned or signed SSL certificates
    • Version 6.1 and above with signed SSL certificates
  • Firefox 3.0 and above
  • Internet Explorer 10 and above

Launching MapR Control System

To use the MapR Control System (MCS), navigate to the host that is running the WebServer in the cluster. MapR Control System access to the cluster is typically via HTTP on port 8080 or via HTTPS on port 8443.You should disable pop-up blockers in your browser to allow MapR to open help links in new browser tabs.

The first time you open the MCS via HTTPS from a new browser, the browser alerts you that the security certificate is unrecognized. This is normal behavior for a new connection. Add an exception in your browser to allow the connection to continue.

Direct Access NFS™

Use Direct Access NFS™ to mount the MapR-FS locally as a directory on a Mac, Linux, or Windows computer.

See Accessing Data with NFS for information.

MapR Client

Use the MapR client to submit MapReduce jobs, to submit YARN applications, and to run MapR CLI, hadoop fs, and hadoop mfs commands. To run MapR CLI commands on a MapR client, establish an ssh session to a node in the cluster.

The method that you use to submit the hadoop commands on Mac and Windows clients is different than the method that is used on Linux machines. See Running Hadoop Commands on a Mac and Windows Client.

Installing the MapR Client

 Before you install the MapR Client, perform the following steps:

  • Verify that the operating system on the machine where you plan to install the MapRClient is supported.
    See the interoperability matrix for the list of operating systems that are compatible with the MapR client.  

  • Verify that the machine where you plan to install the MapR client is not a cluster node. 
    The MapR client is intended for use on a computer that has no other MapR server software installed.

  • Obtain connectivity information and cluster setup requirements. 
    When you use configure.sh to configure the client, you will need to know the following details: 

    • The cluster name. You will need the cluster name when you specify the -N parameter.

    • The IP addresses and ports of the CLDB nodes on the cluster.  You will need this information when you specify the CLDB nodes with the -C parameter.

    • If one or more nodes in the cluster run the ResourceManager, you may need to specify the hostname or IP address for each ResourceManager nodes using the -RM parameter. If the cluster is configured to use zero-configuration failover, do not specify the ResourceManager nodes. If the cluster is not configured to use zero-configuration failover, specify each ResourceManager node.
    • Determine if the cluster is secure. If the cluster is secure, you will need to specify the -secure parameter when you run configure.sh. 
    • If a node in the cluster runs the HistoryServer, note the hostname for the HistoryServer. You must specify each HistoryServer node using the -HS parameter.
  • In the /etc/hosts file of the client machine, add a mapping between the CLDB nodes in the cluster and the IP addresses of those nodes. For example, add the IP address 10.10.82.22 and CLDB node name centos22 on the Mac OSX where you installed the client:

To install the MapR client, obtain the MapR packages for your operating systems at http://package.mapr.com/releases/<version>/<operating system/ and complete the installation steps:

For the Windows MapR client, you must also complete some post installation steps.

Icon

The client nodes also need to have the MapR repositories configured in order to pull the client packages.

 

Installing the MapR Client on CentOS or Red Hat

  1. Remove any previous MapR software. You can use rpm -qa | grep mapr to get a list of installed MapR packages, then type the packages separated by spaces after the rpm -e command. Example:

    rpm -qa | grep mapr
    rpm -e mapr-fileserver mapr-core
  2. Install the MapR client for your target architecture:
    • yum install mapr-client.i386
    • yum install mapr-client.x86_64
  3. Run configure.sh to configure the client. For details about the syntax, parameters, and behavior of configure.sh, see configure.sh.

     Un-secure cluster example:

    In the following example, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -C (uppercase) parameter specifies the CLDB nodes, and the -HS parameter specifies the HistoryServer node:

    /opt/mapr/server/configure.sh -N my.cluster.com -c -C mynode01:7222 -HS mynode02  

     Secure cluster example:

    In the following example, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -secure parameter indicates that the cluster is secure, the -C (uppercase) parameter specifies the CLDB nodes, and the -HS parameter specifies the HistoryServer node: 

    /opt/mapr/server/configure.sh -N my.cluster.com -c -secure -C mynode01:7222 -HS mynode02

  4. To use this client with a secure cluster or clusters, copy the ssl_truststore file from the /opt/mapr/conf directory on the cluster to the /opt/mapr/conf directory on the client. If this client will connect to multiple clusters, merge the ssl_truststore files with the /opt/mapr/server/manageSSLKeys.sh tool. 
    See the Security Guide for details on how to connect to a secure cluster.  

Installing the MapR Client on SUSE

  1. Remove any previous MapR software. You can use rpm -qa | grep mapr to get a list of installed MapR packages, then type the packages separated by spaces after the zypper rm command. 
    Example:
    rpm -qa | grep mapr
    zypper rm mapr-fileserver mapr-core

  2. Run the following command to install the MapR client: zypper install mapr-client
  3. Run configure.sh to configure the client.  For details about the syntax, parameters, and behavior of configure.sh, see configure.sh.

     Un-secure cluster example...

    In the following examples, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -C (uppercase) parameter specifies the CLDB nodes, and -HS parameter specifies the HistoryServer node: 

    /opt/mapr/server/configure.sh -N my.cluster.com -c -C mynode01:7222 -HS mynode02  

     Secure cluster example...

    In the following example, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -secure parameter indicates that the cluster is secure, the -C (uppercase) parameter specifies the CLDB node, and the -HS parameter specifies the HistoryServer node: 

    /opt/mapr/server/configure.sh -N my.cluster.com -c -secure -C mynode01:7222 -HS mynode02

  4. To use this client with a secure cluster or clusters, copy the ssl_truststore file from the /opt/mapr/conf directory on the cluster to the /opt/mapr/conf directory on the client. If this client will connect to multiple clusters, merge the ssl_truststore files with the /opt/mapr/server/manageSSLKeys.sh tool.

 Installing the MapR Client on Ubuntu

  1. Remove any previous MapR software. You can use dpkg --list | grep mapr to get a list of installed MapR packages, then type the packages separated by spaces after the dpkg -r command. 
    Example:
    dpkg -l | grep mapr

    dpkg -r mapr-core mapr-fileserver 
  2. Update your Ubuntu repositories. 
    Example:

    apt-get update
  3. Run the following command to install the MapR client: apt-get install mapr-client
  4. Run configure.sh to configure the client. For details about the syntax, parameters, and behavior of configure.sh, see configure.sh.

     Un-secure cluster example...

    In the following examples, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -C (uppercase) parameter specifies the CLDB node, and -HS specifies the HistoryServer node: 

    /opt/mapr/server/configure.sh -N my.cluster.com -c -C mynode01:7222 -HS mynode02

     Secure cluster example...

    In the following example, the -N parameter specifies the cluster name, the -c (lowercase) parameter specifies a client configuration, the -secure parameter indicates that the cluster is secure, the -C (uppercase) parameter specifies the CLDB node, and the -HS parameter specifies the HistoryServer node: 

    /opt/mapr/server/configure.sh -N my.cluster.com -c -secure -C mynode01:7222 -HS mynode02

  5. To use this client with a secure cluster or clusters, copy the ssl_truststore file from the /opt/mapr/conf directory on the cluster to the /opt/mapr/conf directory on the client. If this client will connect to multiple clusters, merge the ssl_truststore files with the /opt/mapr/server/manageSSLKeys.sh tool.
    See the Security Guide for details on how to connect to a secure cluster.  

Installing the MapR Client on Mac OS X

  1. Create the directory /opt:
    sudo mkdir -p /opt

  2. Download the file for the version that you want to install: http://package.mapr.com/releases/<version>/mac/<mapr-client package name>
  3. Open the Terminal application.
  4. Extract mapr-client-<version>.tar.gz into the /opt directory:
    sudo tar -C /opt -zxf mapr-client-<version>.tar.gz*
  5. Before running configure.sh, make sure that JAVA_HOME is set correctly for the client in the following script:

    /opt/mapr/conf/env.sh

    For example:

    $ export JAVA_HOME=$(/usr/libexec/java_home)
  6. Run configure.sh to configure the client. For details about the syntax, parameters, and behavior of configure.sh, see configure.sh.

     Un-secure cluster example...

    In the following examples, the -N parameter specifies the cluster name, the -c (lowercase) option specifies a client configuration, the -C (uppercase) parameter specifies the CLDB nodes, and -HS specifies the HistoryServer node: 

    sudo /opt/mapr/server/configure.sh -N my.cluster.com -c -C mynode01:7222 -HS nodeA

     Secure cluster example...
    In the following example, the -N parameter specifies the cluster name, the -C (uppercase) parameter specifies the CLDB nodes, the -c (lowercase) option specifies a client configuration, and -secure indicates that the cluster is secure: 
    sudo /opt/mapr/server/configure.sh -N my.cluster.com -c -secure -C mynode01:7222
  7. To use this client with a secure cluster or clusters, copy the ssl_truststore file from the /opt/mapr/conf directory on the cluster to the /opt/mapr/conf directory on the client. If this client will connect to multiple clusters, merge the ssl_truststore files with the /opt/mapr/server/manageSSLKeys.sh tool.
    See the Security Guide for details on how to connect to a secure cluster.  

Limitation: Under OS X, the getgroups command returns a maximum of 16 groups for a user. If the Mac OS user for which you are installing the MapR Client attempts to read or write to a MapR-FS resource as a member of a group that was not included in the list of 16 groups returned by getgroups, file permission errors may result.

Installing the MapR Client on Windows

  1. Make sure Java is installed on the computer and that the JAVA_HOME environment variable is set correctly. 

    Icon

    The path that you set for the JAVA_HOME environment variable should not include spaces.

  2. Create the directory \opt\mapr on your c: drive (or another hard drive of your choosing).
    You can use Windows Explorer or type the following at the command prompt:
    mkdir c:\opt\mapr
  3. Set the MAPR_HOME environment variable to c:\opt\mapr 
  4. Open the command line.
  5. Use the following command to navigate to MAPR_HOME:
    cd %MAPR_HOME%
  6. Download the file, for the version that you want to install, into MAPR_HOMEhttp://package.mapr.com/releases//windows/ package name>
  7. Extract the archive by right-clicking on the file and selecting Extract All...
  8. From the command line, run configure.bat to configure the client. For details about the syntax, parameters, and behavior of configure.bat, see configure.sh.

     Un-secure cluster example...

    In the following examples, the -N parameter specifies the cluster name, the -c (lowercase) option specifies a client configuration, the -C (uppercase) parameter specifies the CLDB nodes, and -HS specifies the HistoryServer node: 

    server\configure.bat -N my.cluster.com -c -C mynode01:7222 -HS nodeA  

     Secure cluster example...
    In the following example, the -N parameter specifies the cluster name, the -C (uppercase) parameter specifies the CLDB nodes, the -c (lowercase) option specifies a client configuration, and -secure indicates that the cluster is secure:
    server\configure.bat -N my.cluster.com -c -secure -C mynode01:7222
  9. To use this client with a secure cluster or clusters, copy the ssl_truststore file from the /opt/mapr/conf directory on the cluster to the c:\opt\mapr\conf directory on the client. If this client will connect to multiple clusters, merge the ssl_truststore files with the c:\opt\mapr\server\manageSSLKeys.bat tool
    See the Security Guide for details on how to connect to a secure cluster.   

Compatibility with Network Address Translation (NAT) Adapters

In VM environments, the MapR Client on Windows works with a single NAT virtual adapter as long as it is the only virtual adapter configured for the VM. If you want to use more than one adapter, you must use other types of virtual adapters. If you use multiple NAT adapters in your VM environment, your jobs and file system operations will fail.

Using the Windows Client

This section contains the following topics about using the MapR client on Windows: 

Configuring MapR Client User on Windows

Before running jobs or applications on the Windows Client, configure the core-site.xml with the UID, GID, and user name of the cluster user that will be used to access the cluster.

Complete the following steps:

  1. Obtain the UID and GID that has been set up for your user account.
    To determine the correct UID and GID values for your username, log into a cluster node and type the id command. In the following example, the UID is 1000 and the GID is 2000:

  2. Add the following parameters to the core-site.xml files that correspond to the version of the hadoop commands that you plan to run:

    Icon

    You must use the numeric values for UID and GID, not the text names.

    Icon

    When wire-level security is implemented on Windows, spoofing is not supported. This greatly increases the security of the cluster, but the core-site.xml file settings above then have no effect.

    The location of the core-site.xml file(s) that you need to edit is based on the type of job or applications that you will run from this client machine:

    Job or Application Typecore-site.xml Location
    MapReduce v1 jobs%MAPR_HOME%\hadoop\hadoop-0.20.0\conf\core-site.xml
    YARN applications
    (MapReduce v2 or other applications that run on YARN)
    %MAPR_HOME%\hadoop\hadoop-2.x.x\etc\hadoop\core-site.xml
    Icon

    If you plan to run both MapReduce v1 jobs and MapReduce v2 applicatons, you need to update both core-site.xml files with the user parameters.

Configure the MapR Windows Client to Submit MapReduce V2 Applications

Before running any MapReduce v2 applications from a Windows client, add the following property to the mapred-site.xml file:

The mapred-site.xml file for MapReduce v2 applications is located in the following directory:  %MAPR_HOME%\hadoop\hadoop-2.x.x\etc\hadoop\mapred-site.xml

Running Hadoop Commands on a Mac and Windows Client

When you run hadoop commands on the Mac and Windows client,  the location from which you run the command differs based on the command version that you require. For example, use the Hadoop 1 version of the hadoop jar command to run MapReduce v1 jobs and use the Hadoop 2 version to run MapReduce v2 applications. 

To run the..Run the command from this location:
Hadoop 1 version of the command

On Windows: %MAPR_HOME%\hadoop\hadoop-0.20.0\bin

On Mac: /opt/mapr/hadoop/hadoop-0.20.2/bin

Hadoop 2 version of the command

On Windows: %MAPR_HOME%\hadoop\hadoop-2.x.x\bin

On Mac: /opt/mapr/hadoop/hadoop-2.x.x/bin

On Linux installations, the installer creates symlinks to the hadoop directories by default. On Mac, you can create the symlinks.  Once the symlinks are created, you can specify the version of the hadoop command as mentioned in the Hadoop command documentation

Notes Specific to Windows

Icon
  • The user that runs hadoop commands from the hadoop 2 directory cannot have a space or a hyphen (-)  in the username.
  • The native Hadoop library is not present on Windows. Therefore, the hadoop fs -getmerge command is not available.

Create Symlinks to Hadoop Directories for the Mac Client

Perform the following steps to create hadoop1 and hadoop2 symlinks in the usr/local/bin directory for a MapR client on Mac OS X:

  1. Run the following commands as root to create the symlinks:

    Icon
    In the command above, replace hadoop-2.x.x with the actual hadoop 2 version number that you have installed.
  2. Add the Hadoop binaries to the PATH environment variable.
    For example, add the following text to the user login shell script such as ~/.bashrc: 

    Icon
    In the text above, replace hadoop-2.x.x with the actual hadoop 2 version number that you have installed.

Now, you can run hadoop commands using the hadoop1 and hadoop2 keywords, as mentioned in the Hadoop command documentation.

 

 

 

  • No labels