On-premise Installation

Prerequisites

Supported Operating Systems

• CentOS - Versions 6.6+, 7.3
• Red Hat Enterprise Linux - Versions 6.6+, 7.3
• Ubuntu - Version 16.04 (supported for HDInsight only)
• Debian - 8.1 (supported for DataProc only)

Supported Hadoop Distributions

• HDP - Versions 2.5.5, 2.6.4, 3.x
• MAPR - Version 6.0.1
• Cloudera - Version 5.13
• Azure - HDI 3.6
• GCP - 1.2 (Unsecured), 1.3 (Secured) Dataproc
• EMR - Version 5.17.0

Installation Procedure

The installation logs are available in <path_to_Infoworks_home>/iw-installer/logs/installer.log.

Perform the following:

Download and Extract Installer

• Download the installer tar ball: wget <link-to-download>
• Extract the installer: tar -xf deploy_<version_number>.tar.gz
• Navigate to installer directory: cd iw-installer

Configure Installation

• Run the following command: ./configure_install.sh

Enter the details for each prompt:

• Hadoop distro name and installation path (If not auto-detected)
• Infoworks user
• Infoworks user group
• Infoworks installation path
• Infoworks HDFS home (path of home folder for Infoworks artifacts)
• Hive schema for Infoworks sample data
• IP address for accessing Infoworks UI (when in doubt use the FQDN of the Infoworks host)
• HiveServer2 thrift server hostname
• Hive user name
• Hive user password

If Hadoop distro is Cloudera (CDH):

• Impala hostname
• Impala port number
• Impala user name
• Impala password
• Is Impala Kerberized?

If Impala is Kerberized:

• Kerberos Realm
• Kerberos host FQDN

If Hadoop distro is GCP:

• Managed Mongo URL
• Are infoworks directories already extracted in IW_HOME?

Run Installation

• Install Infoworks: ./install.sh -v <version_number>

NOTE: For machines without certificate setup, --certificate-check parameter can be entered as false as described in the following syntax: ./install.sh -v <version_number> --certificate-check <true/false>. The default value is true. If you set it to false, this performs insecure request calls. This is not a recommended setup.

NOTE:

For HDP, CentOS/RHEL6, replace <version_number> with 2.9.0-hdp-rhel6

For HDP, CentOS/RHEL7, replace <version_number> with 2.9.0-hdp-rhel7

For MapR or Cloudera, CentOS/RHEL6, replace <version_number> with 2.9.0-rhel6

For MapR or Cloudera, CentOS/RHEL7, replace <version_number> with 2.9.0-rhel7

For Azure, replace <version_number>with 2.9.0-azure

For GCP, replace <version_number>with 2.9.0-gcp

For EMR, replace <version_number>with 2.9.0-emr

Post Installation

If the target machine is Kerberos enabled, performed the following post installation steps:

• Go to <IW_HOME>/conf/conf.properties
• Edit the Kerberos security settings as follows (ensure these settings are uncommented):

xxxxxxxxxx

#Security Settings

#iw_security_impersonate=true

#iw_security_kerberos_enabled=true

#iw_security_kerberos_default_principal=INFOWORKS-USER@DOMAIN.COM

#iw_security_kerberos_default_keytab_file=/path/to/infoworks-user.keytab

#iw_security_kerberos_hiveserver_principal=hive/HIVEHOST@DOMAIN.COM)

• Restart the Infoworks services.

NOTE: Kerberos tickets are renewed before running all the Infoworks DataFoundry jobs. Infoworks DataFoundry platform supports single Kerberos principal for a Kerberized cluster. Hence, all Infoworks DataFoundry jobs work using the same Kerberos principal, which must have access to all the artifacts in Hive, Spark, and HDFS.