Required settings

The first part of bdd.conf contains required settings. You must update these with information specific to your system, or the installer could fail.

Must Set

This section contains blank settings that you must provide values for. If you don't set these, the installation will fail.

Configuration property Description
ORACLE_HOME The path to the BDD root directory, where BDD will be installed on each node in the cluster. This directory must not exist, and its parent directories' permissions must be set to either 755 or 775.

Note that this is different from the ORACLE_HOME environment variable required by the Studio database.

Important: You must ensure that the installer can create this directory on all nodes that will host BDD components, including Hadoop nodes that will host Data Processing.

On the install machine and all other nodes that will host WebLogic Server, this directory must contain at least 6GB of free space. Nodes that will host the Dgraph require 1GB of free space, and those that will host Data Processing require 2GB.

ORACLE_INV_PTR The absolute path to the Oracle inventory pointer file, which the installer will create. This file can't be located in the ORACLE_HOME directory.

If you have any other Oracle software products installed, this file will already exist. Update this property to point to it.

INSTALLER_PATH Optional. The absolute path to the installation source directory. This must contain at least 10GB of free space.

If you don't set this property, you can either set the INSTALLER_PATH environment variable or specify the path at runtime. For more information, see The BDD installer.

DGRAPH_INDEX_DIR The absolute path to the Dgraph databases. This shouldn't be located under ORACLE_HOME, or it will be deleted.

The script will create this directory if it doesn't currently exist. If you're installing with existing databases, set this property to their parent directory.

If you're installing on a CDH cluster with HDFS data at rest encryption enabled and you want to store your databases on HDFS, be sure that this is in an encryption zone.

HADOOP_UI_HOST The name of the server hosting your Hadoop manager (Cloudera Manager or Ambari).
STUDIO_JDBC_URL The JDBC URL for the Studio database.
There are three templates for this property. Copy the template that corresponds to your database type to STUDIO_JDBC_URL and update the URL to point to your database.
  • If you have a MySQL database, use the first template and update the URL as follows: 
    jdbc:mysql://<database hostname>:<port number>/<database name>?useUnicode=true&characterEncoding=UTF-8&useFastDateParsing=false
  • If you have an Oracle database, use the first template and update the URL as follows: 
    jdbc:oracle:thin:@<database hostname>:<port number>:<database SID>
  • If you're not installing on a production environment and want the installer to create a Hypersonic database for you, use the third template. The script will create the database for you in the location defined by the URL.
Note: BDD doesn't currently support database migration. After deployment, the only ways to change to a different database are to reconfigure the database itself or reinstall BDD.

General

This section configures settings relevant to all components and the installation process itself.

Configuration property Description
INSTALL_TYPE Determines the installation type according to your hardware and Hadoop distribution. Set this to one of the following:
  • CDH
  • HW

Note that this document doesn't cover Oracle Big Data Appliance (BDA) or Oracle Public Cloud (OPC) installations. If you want to install on the Big Data Appliance, see the Oracle Big Data Appliance Owner's Guide Release 4 (4.5) and the corresponding MOS note.

CLUSTER_MODE Determines whether you're installing on a single machine or a cluster. Use TRUE if you're installing on a cluster, and FALSE if you're installing on a single machine.
Note: If you're installing on a single machine, see Single-Node Installation.
JAVA_HOME The absolute path to the JDK install directory. This must be the same on all BDD servers and should have the same value as the $JAVA_HOME environment variable.

If you have multiple versions of the JDK installed, be sure that this points to the correct one.

TEMP_FOLDER_PATH The temporary directory used on each node during the installation. This must point to an existing directory on all BDD nodes.

On the install machine and all other WebLogic and Dgraph nodes, this directory must contain at least 10GB of free space. Data Processing nodes require 3GB of free space.

CDH/HDP

This section contains properties related to Hadoop. The installer uses these properties to query the Hadoop manager (Cloudera Manager or Ambari) for information about the Hadoop components, such as the URIs and names of their host servers.

Configuration property Description and possible settings
HADOOP_UI_PORT The port number of the server running the Hadoop manager.
HADOOP_UI_CLUSTER_NAME The name of your Hadoop cluster, which is listed in the manager. Be sure to replace any spaces in the cluster name with %20.
HUE_URI HDP only. The hostname and port of the node running Hue, in the format <hostname>:<port>.
HADOOP_CLIENT_LIB_PATHS A comma-separated list of the absolute paths to the Hadoop client libraries.
Note: You only need to set this property before installing if you have HDP. For CDH, the installer will download the required libraries and set this property automatically. Note that this requires an internet connection. If the script is unable to download the libraries, it will fail; see Failure to download the Hadoop client libraries for instructions on solving this issue.

To set this property, copy the template for your Hadoop distribution to HADOOP_CLIENT_LIB_PATHS and update the paths to point to the client libraries you copied to the install machine.

Don't change the order of the paths in the list as they must be specified as they appear.

HADOOP_CERTIFICATION_PATH Only required for Hadoop clusters with TLS/SSL enabled. The absolute path to the directory on the install machine where you put the certificates for HDFS, YARN, Hive, and the KMS.

Kerberos

This section configures Kerberos for BDD.

Note: You only need to modify these properties if you want to enable Kerberos.
Configuration property Description and possible settings
ENABLE_KERBEROS Enables Kerberos in the BDD cluster. If Kerberos is installed on your cluster and you want BDD to integrate with it, set this value to TRUE; if not, set it to FALSE.
KERBEROS_PRINCIPAL The name of the BDD principal. This should include the name of your domain; for example, bdd-service@EXAMPLE.COM.

This property is only required if ENABLE_KERBEROS is set to TRUE.

KERBEROS_KEYTAB_PATH The absolute path to the BDD keytab file on the install machine.

The installer will rename this to bdd.keytab and copy it to $BDD_HOME/common/kerberos/ on all BDD nodes.

This property is only required if ENABLE_KERBEROS is set to TRUE.

KRB5_CONF_PATH The absolute path to the krb5.conf file on the install machine. The installer will copy this to /etc on all BDD nodes.

This property is only required if ENABLE_KERBEROS is set to TRUE.

WebLogic (BDD Server)

This section configures the WebLogic Server, including the Admin Server and all Managed Servers.

Configuration property Description and possible settings
ADMIN_SERVER The hostname of the install machine, which will become the Admin Server.

If you leave this blank, it will default to the hostname of the machine you're on.

MANAGED_SERVERS A comma-separated list of the Managed Server hostnames (the servers that will run WebLogic, Studio, and the Dgraph Gateway). This list must include the Admin Server and can't contain duplicate values.
Note: If you define more than one Managed Server, you must set up a load balancer in front of them after installing. For more information, see Configuring load balancing.

Dgraph and HDFS Agent

This section configures the Dgraph and the HDFS Agent.

Configuration property Description and possible settings
DGRAPH_SERVERS A comma-separated list of the hostnames of the nodes that will run the Dgraph and the Dgraph HDFS Agent.

This list can't contain duplicate values. If you plan on storing your databases on HDFS, these must be HDFS DataNodes. For best performance, there shouldn't be any other Hadoop services running on these nodes, especially Spark.

DGRAPH_THREADS The number of threads the Dgraph starts with. This should be at least 2. The exact number depends on the other services running on the machine:
  • For machines running only the Dgraph, the number of threads should be equal to the number of cores on the machine.
  • For machines running the Dgraph and other BDD components, the number of threads should be the number of cores minus two. For example, a quad-core machine should have two threads.
  • For HDFS nodes running the Dgraph, the number of threads should be the number of CPU cores minus the number required for the Hadoop services. For example, a quad-core machine running Hadoop services that require two cores should have two threads.

If you leave this property blank, it will default to the number of CPU cores minus two.

Be sure that the number you use is in compliance with the licensing agreement.

DGRAPH_CACHE The size of the Dgraph cache, in MB. Only specify the number; don't include MB.

If you leave this property blank, it will default to either 50% of the node's available RAM or the total mount of free memory minus 2GB (whichever is larger).

Oracle recommends allocating at least 50% of the node's available RAM to the Dgraph cache. If you later find that queries are getting cancelled because there isn't enough available memory to process them, experiment with gradually decreasing this amount.

ZOOKEEPER_INDEX The index of the Dgraph cluster in the ZooKeeper ensemble, which ZooKeeper uses to identify it.

Data Processing

This section configures Data Processing and the Hive Table Detector.

Configuration property Description and possible settings
HDFS_DP_USER_DIR The location within the HDFS /user directory that stores the Avro files created when Studio users export data. The installer will create this directory if it doesn't already exist. The name of this directory must not include spaces or slashes (/).
YARN_QUEUE The YARN queue Data Processing jobs are submitted to.
HIVE_DATABASE_NAME The name of the Hive database that stores the source data for Studio data sets.

The default value is default. This is the same as the default value of DETECTOR_HIVE_DATABASE, which is used by the Hive Table Detector. It is possible to use different databases for these properties, but it is recommended that you start with one for a first time installation.

SPARK_ON_YARN_JAR The absolute path to the Spark on YARN JAR on your Hadoop nodes. This will be added to the CLI classpath.
There are two templates for this property. Copy the value of the template that corresponds to your Hadoop distribution to SPARK_ON_YARN_JAR and update its value as follows:
  • If you have CDH, use the first template. This should be the absolute path to spark-assembly.jar.
  • If you have HDP, use the second template. This should be the absolute paths to hive-exec.jar and spark-assembly.jar, separated by a colon: 
    <path/to/hive-exec.jar>:<path/to/spark-assembly.jar>

Micro Service

This section configures the Transform Service.

Configuration property Description and possible settings
TRANSFORM_SERVICE_SERVERS A comma-separated list of the Transform Service nodes. For best performance, these should all be Managed Servers. In particular, they shouldn't be Dgraph nodes, as both the Dgraph and the Transform Service require a lot of memory.
Note: If you define more than one Managed Server, you must set up a load balancer in front of them after installing. For more information, see Configuring load balancing.
TRANSFORM_SERVICE_PORT The port the Transform Service listens on for requests from Studio.