This page explains how to install Trafodion on a cluster. The instructions apply to the single-script installer and the latest Trafodion release, which is Release 1.1.

Preparing Your PC

If you are using a Windows PC, additional PC software may be needed throughout the installation process. It is recommended that you preinstall the PC software that you may need before continuing with the Trafodion installation.

Install this PC software:

  • putty and puttygen (download from the PuTTY website)
  • VNC client (download from http://www.realvnc.com)
  • Firefox or Chrome browser
  • SFTP client to transfer files from your PC to the server: WinSCP or FileZilla

Preparing Your Cluster Environment

Before installing Trafodion, see these sections to understand the requirements and to prepare for installation:

 

Cluster Requirements

These configuration settings have been tested and are known to work with the Trafodion installation:

Hardware platformx86-64
Operating system
  • RedHat Enterprise Linux, 6.x kernel (64 bit)
  • CentOS, 6.x kernel (64 bit)
User IDsA user ID with passwordless sudo access. See Obtaining Sudo Access and Passwordless SSH.
Cluster sizeA cluster consisting of 1 to N nodes. There is currently no upper limit. Two nodes is the recommended minimum, but a single node cluster should work.
Disk spaceMinimum of 20 GB. See Checking the Disk Space (Cloudera Only).
MemoryMinimal 1 GB * number of connectivity servers (MXOSRVR processes) configured per cluster.

 

Checking the Disk Space (Cloudera Only) 

Before installing Cloudera, a minimum of 20 GB is needed to support the database. The default installation location is "/var/lib/cloudera-scm-server-db". To check the size available to /var, start a putty session, or start a VNC terminal window on the node in your cluster where Cloudera will be installed. To execute the following commands, you will need either root or sudo access.

Confirm there is a minimum of 20 GB available to /var.

cd "/var"
df -hP    

If there is insufficient space available in /var, a possible solution is to provide a soft link to another drive for your Cloudera database. Locate a drive that does have sufficient space.

cd <new drive>  (e.g. cd /DATA)
mkdir cloudera-scm-server-db
chmod 777 cloudera-scm-server-db
cd /var/lib
ln -s <new drive>/cloudera-scm-server-db .

 

If /var is a subdirectory in your cluster's root filesystem, the Cloudera database should have sufficient space available.

 

If Cloudera has already been installed and it is showing red for log directories, it means that Cloudera was installed using /var/lib and may not have a large enough /var file system. In this case, there is an unsupported script that can be used to move the directories. Please see the clouderaMoveDB.sh script in the installer/tools directory, which is created when the installer tar.gz file is untarred. Execute clouderaMoveDB.sh without syntax to display help information.

Obtaining Sudo Access and Passwordless SSH

The Trafodion installation requires a user ID with these attributes:

  • sudo access
  • passwordless ssh to all nodes on the cluster

 

You may need to request permission from your cluster management team to obtain this type of access.

 

The following example shows how to set up your user ID to have "passwordless ssh" abilities.

echo -e 'y\n' | ssh-keygen -t rsa -N "" -f $HOME/.ssh/id_rsa
cat $HOME/.ssh/id_rsa.pub >> $HOME/.ssh/authorized_keys
echo localhost $(cat /etc/ssh/ssh_host_rsa_key.pub) >> $HOME/.ssh/known_hosts
echo "NoHostAuthenticationForLocalhost=yes" >> $HOME/.ssh/config
chmod 600 $HOME/.ssh/config

After running these commands, you will need to copy the contents of the public key file, $HOME/.ssh/id_rsa.pub, and append those contents to each additional node's $HOME/.ssh/authorized_keys file. You will also need to copy your private $HOME/.ssh/id_rsa file from the head node to the other nodes' $HOME/.ssh directory, and secure it as private to yourself (chmod 700), as you do not want someone else pretending to be you.

 

Disabling requiretty

You will need to disable requiretty so that sudo commands can be run from inside the installation scripts.

To make sure that the requiretty option is NOT being used, in the /etc/sudoers file, comment out the "Defaults requiretty" setting.

 

Configuring an LDAP Identity Store

If you plan to enable authentication in Trafodion, you will need to have an LDAP identity store available to perform authentication. The Trafodion installer will prompt you to set up an authentication configuration file that points to an LDAP server (or servers), which will enable security (that is, authentication and authorization) in the Trafodion system. If you wish to manually set up the authentication configuration file and enable security, see Enabling Security Features in Trafodion.

Installing a Supported Hadoop Distribution

Trafodion Release 1.1 is compatible with the following distributions:

 

DistributionVersionHBase VersionInstallationAdditional Information
Cloudera Distribution Including Apache Hadoop (CDH)CDH 5.3 (or 5.2)0.98.6To install CDH 5.3, see the CDH 5.3.x Installation and Upgrade Guide. Before installing CDH, please see the note below and review theMandatory Hadoop Services and Settings.

NOTE:

  • Select Cloudera Standard edition.
  • Install using packages. Do not install using tarballs and parcels. Trafodion does not currently support tarball/parcel installation.
  • CDH 5.2 is supported but has not been officially tested with Trafodion Release 1.1.
  • If you are installing Trafodion Release 1.0, it supports CDH 5.1.n and HBase 0.98.1.
Mandatory Hadoop Services and Settings
Hortonworks Data Platform (HDP)HDP 2.20.98.4To install HDP 2.2, see the Ambari 1.7.0 Install Guide for a first-time installation or the Ambari 1.7.0 Upgrade Guide to upgrade to HDP 2.2. Before installing HDP, please review the Mandatory Hadoop Services and Settings.

NOTE: If you are installing Trafodion Release 1.0, it supports HDP 2.1 and HBase 0.98.0.

Mandatory Hadoop Services and Settings

 

Mandatory Hadoop Services and Settings

IMPORTANT: Before installing a Hadoop distribution, please review this list of mandatory services and settings:

  • HDFS
  • MapReduce
  • ZooKeeper
  • HBase
  • Hive
  • Nagios (Hortonworks)
  • Ganglia (Hortonworks)
  • Embedded Databases (Cloudera)

Please make sure that those services and settings are selected during installation.

IMPORTANT: Trafodion does not support a Hadoop server (for example, Cloudera or Hortonworks) running on a node that is not part of the Trafodion cluster. You must specify the Hadoop nodes to be part of the Trafodion list of nodes during the installation of Trafodion (using the --nodes parameter). The Trafodion installer needs to run from the same node where you installed the Hadoop distribution (Cloudera or Hortonworks).

User IDs and Passwords

For your reference, this table lists the user IDs and passwords that you will use during installation.

NOTE: On a physical cluster, you will use two user IDs: your user ID and trafodion.

LogonUser IDPasswordDescription
Cloudera Manager Web GUI logonadmin(default)admin(default)After installing Cloudera, you will be instructed to log on to the Cloudera Manager Web GUI. Use the default user ID and password (adminadmin). If you already had Cloudera installed, please use your previously defined user ID and password.
Ambari Web GUI logonadmin(default)admin(default)After downloading and setting up the Ambari repository, you will be instructed to log on to the Ambari Web GUI. Use the default user ID and password (adminadmin). If you already had Ambari installed, please use your previously defined user ID and password.
User ID with sudo access<sudo-username><password>In the installation steps, you may be instructed to use "sudo" or "sudo userid" access. You will be using your user ID, which has been enabled with "sudo" access and passwordless ssh to all nodes of the cluster.
Trafodion logontrafodiontraf123(default)This user ID is automatically created for you by the Trafodion installer when Trafodion is installed. Do not create this user ID manually.
VNC server logontrafodiontraf123(default)After installing Trafodion, you will be instructed to log on to the VNC server at port 1. Use the trafodion user ID and password.

 

Installing Trafodion

NOTE: If you are upgrading to a newer version of Trafodion, please see Upgrading Trafodion before proceeding.

At a high level, the following instructions consist of a single step:

 

CommandUser IDComments
trafodion_install<sudo-username>This installs necessary RPMs, creates the trafodion user ID, sets up passwordless ssh for the trafodion user ID, copies the Trafodion distribution files across the cluster, generates startup files, and starts Trafodion, including Database Connectivity Services (DCS) (the ODBC/JDBC server), among other things.

 

IMPORTANT: Those Trafodion commands should all be done from the first node of the cluster. Do not run them from a machine that is not part of the Trafodion cluster. Trafodion must be installed on all nodes that host an HBase RegionServer (that is, where a supported Hadoop distribution is installed). For example, if Cloudera or Hortonworks is installed on nodes n001, n002, and n003, then you must install Trafodion on those same nodes (--nodes "n001 n002 n003").

 

    1. If you have not done so already, download the latest tar files from downloads.trafodion.org.

      NOTE: The Trafodion installer operates on only the Trafodion binaries downloaded from downloads.trafodion.org. The installer does not work on builds of the Trafodion source code from the GitHub repository. To learn more about the process of contributing to the source code, please see the instructions in Contributing to the Software.

    2. If you are upgrading to a newer version of Trafodion, copy the installation files for the previous version to a directory outside of $HOME/trafodion_downloads. For example:
      mkdir $HOME/trafodion_downloads_v1.0
      cp –r $HOME/trafodion_downloads/* $HOME/trafodion_downloads_v1.0
      rm –rf $HOME/trafodion_downloads/*
      

 

 

    1. Create the trafodion_downloads directory if you have not done so already and place the downloaded trafodion-1.1.<n>.tar.gz and installer-1.1.<n>_v<num>.tar.gz tar files into it:
      mkdir $HOME/trafodion_downloads
      mv <your-download-path>/trafodion-1.1.<n>.tar.gz $HOME/trafodion_downloads
      mv <your-download-path>/installer-1.1.<n>_v<num>.tar.gz $HOME/trafodion_downloads
      cd $HOME/trafodion_downloads
      

 

 

    1. Untar the downloaded installer file:
      tar -xzf installer-1.1.<n>_v<num>.tar.gz
      

 

 

    1. Change to the installer directory:
      cd installer
      

 

 

    1. Run the trafodion_install script. There are no parameters for it. You will be prompted for all needed information.
      ./trafodion_install
      

      NOTE: The trafodion_install script automatically invokes the Trafodion Scanner to check the cluster for attributes that commonly cause installation issues, runtime performance problems, and so on. Also, various system information is collected for future reference, such as to aid with troubleshooting. The installation will abort if any configuration errors are discovered. You will need to correct such errors before re-running the trafodion_installscript. Please refer to the Trafodion Scanner page for details.

    2. Type "accept" to accept the license agreement when prompted.
    3. Provide the requested information according to the prompts.

      NOTE: If you choose not to start Trafodion after the installation (that is, if you enter N for Start Trafodion after install (Y/N)), you will need to manually start and initialize Trafodion after trafodion_install completes. See step 11.

    4. Wait for trafodion_install to complete. When it completes successfully, you will see this message:
      ***INFO: Installation completed successfully.
      

 

 

    1. Once trafodion_install completes without error, your Trafodion system should be up and running. Start a new VNC session using port :1. Log in as the trafodion user with the password for the VNC, which is traf123 by default unless you changed it.

 

 

  1. If you chose not to start Trafodion after the installation, start and initialize Trafodion as follows:
    cds
    sqstart
    
    [trafodion@n001 ~]$ sqlci
    Trafodion Conversational Interface 1.1.0
    (c) Copyright 2015 Hewlett-Packard Development Company, LP.
    >>initialize trafodion;
    
  2. Perform a quick sanity check using the Trafodion Conversational Interface (sqlci). Create a table with a couple of records. For example:
    [trafodion@n001 ~]$ sqlci
    Trafodion Conversational Interface 1.1.0
    (c) Copyright 2015 Hewlett-Packard Development Company, LP.
    >>create table test1 (f1 int, f2 int);
    
    --- SQL operation complete.
    >>insert into test1 values(1,1);
    
    --- 1 row(s) inserted.
    >>insert into test1 values(2,2);
    
    --- 1 row(s) inserted.
    >>select * from test1;
    
    F1            F2
    -----------   -----------
              1             1
              2             2
    
    --- 2 row(s) selected.
    >>get tables;
    
    Tables in Schema TRAFODION.SEABASE
    ==================================
    
    TEST1
    
    --- SQL operation complete.
    >>exit;
    

    Those SQL commands should run successfully.

  3. Download and install the Trafodion JDBC and/or ODBC drivers on your client workstation to be able to connect to Trafodion from a client application. For instructions, see the Trafodion Client Installation Guide (pdf, 1.12 MB). That guide explains how to install the JDBC and ODBC drivers, how to connect to Trafodion, and how to run sample programs to test the connection.

 

NOTE:

 

  • To troubleshoot issues during or after installation, see Troubleshooting the Installation.
  • After installing Trafodion, see Configuration for guidelines on how to configure your Trafodion environment for optimal performance.
  • If you did not enable security during installation but want to restrict users or restrict access only to registered users, see Enabling Security Features in Trafodion.
  • After successful installation, Trafodion should already be started and ready to use. If you need to stop and restart Trafodion, such as to install a new build, see Starting and Stopping Trafodion.
  • If you want to move data from another relational database management system (RDBMS) to Trafodion, see the guidelines in Data Loading.

 

Upgrading Trafodion

 

Follow these steps if you have an older version of Trafodion that you want to upgrade:

 

  1. Take the Trafodion system offline. See Stopping Trafodion.
  2. If you have important data that you want to retain, back up the metadata and user objects using the Trafodion backup/restore utility. See Backup and Restore.
  3. Install the latest Trafodion software. See Installing Trafodion. The trafodion_install script automatically detects the previous version of Trafodion and performs an upgrade operation, including initializing Trafodion, upgrading the metadata, and restarting all processes.
  4. If something goes wrong during the installation, restore your data from the backed up objects and repeat the previous step.

 

NOTE: If you are doing an upgrade outside of the installation environment, you will need to restart your services and processes, such as sqlci, the DCS server, and the qms servers. If you do not restart those services and processes, your system will not work properly.


  • No labels