Db2 Big SQL configuration utility
A utility is available to configure the Db2® Big SQL service on Cloudera Data Platform (CDP).
Usage notes
The Db2 Big SQL configuration utility bigsql-config is located in the /usr/ibmpacks/IBM-Big_SQL/7.1.0.0/bigsql-cli/ directory after the Db2 Big SQL top-level package is installed.
You must run the bigsql-config utility on the Db2 Big SQL Head node.
A superuser (root or sudo user) or the bigsql user can run the utility. If you
run the utility as the bigsql user, the utility shows only the options that do
not require superuser privilege. For options that only a superuser can run, you must log on as a
superuser. Run the utility with the -help option to see the options that are
available for each user.
Run the bigsql-config utility before you install the Db2 Big SQL service to set installation parameter values, such as Cloudera Manager connection information.
You also run the bigsql-config utility to configure security and whether you want to enable the integration of Db2 Big SQL with HBase. You can set these configuration options before or after you install Db2 Big SQL. After Db2 Big SQL is installed, you can run the utility directly from the shell after you log on to the bigsql account.
The following table describes the parameters that you can configure.
| Configuration parameter | Description | Default value |
|---|---|---|
| BIGSQL_USER | Db2 Big SQL service user name. | bigsql |
| BIGSQL_USER_ID | Db2 Big SQL service numeric uid. | 2824 |
| BIGSQL_GROUP | Db2 Big SQL service group name. | hadoop |
| HIVE_USER | Hive user name. | hive |
| HIVE_GRP | Hive group name. | hive |
| HDFS_USER | HDFS user name. Must be a valid HDFS superuser. |
hdfs |
| HBASE_USER | HBase user name (if HBase is installed). | hbase |
| HADOOP_GRP | Hadoop user group name. | hadoop |
| USERS_GRP | Proxy user group. | users |
| SYSADMGRP_ID | Db2 Big SQL service group numeric gid. | 43210 |
| SUDO_SSH_USER | Sudo SSH user. | root |
| SECURITY_TYPE | User authentication type (ldap, kerberos, ldap_kerberos). The default value is os, which means operating system local users. |
os |
| CUSTOM_PAM_FILE | If Pluggable Authentication Modules (PAM) is enabled, the PAM file location. This file must exist on all Db2 Big SQL nodes. | |
| DO_HIVE_WAREHOUSE_SETUP | Specifies whether HDFS file Access Control Lists (ACLs) are created to grant permissions to
the Db2 Big SQL service user across the Hive warehouse. Notes: If you have a very large number of files and directories, this
operation might take a long time to complete, and potentially exceed the timeout of the
install.
If you have Ranger installed on the cluster, a Ranger policy is preferable to HDFS ACLs. |
True |
| DO_NODE_MANAGER_SETUP | Specifies whether to create the Db2 Big SQL user on
NodeManager nodes. Note: YARN jobs that are run by LOAD and ACID table compaction need the Db2 Big SQL user on NodeManager nodes. Change
DO_NODE_MANAGER_SETUP=True if you plan to use these statements.
|
False |
| ALLOW_TIME_OUT | Specifies whether to allow Db2 Big SQL worker nodes to time out during the installation process. | False |
| ALLOW_FAILURES | Specifies whether to allow Db2 Big SQL worker nodes to fail during the installation process. | False |
| REQUIRED_WORKER_PERCENTAGE | During the installation process, the minimum percentage of Db2 Big SQL workers that must successfully complete a step before continuing with the next step in the process. | 100 |
| TIME_OUT_LIMIT | The maximum timeout limit, in minutes, after the last Db2 Big SQL worker completes the current step in the installation process. | 5 |
| VAR_DIR | Directory where Db2 Big SQL puts files that are large or can potentially grow, such as log files. | /var/ibm/bigsql |
| DBPATH | Location of the Db2 Big SQL metastore. If you are using GPFS, keep the database path local across all Db2 Big SQL nodes. |
/var/ibm/bigsql/database |
| DATA_DIRECTORIES | Location of the Db2 Big SQL data directories. Best practice is to spread the directories across all of the same physical disks that are assigned to store Hadoop data. To specify multiple paths on multiple physical disks, separate each path with a comma. If the use of local Db2 tables is anticipated, best practice is to ensure that the Db2 Big SQL head node also has physical disks mounted at the paths that are specified here. You can have the directories on GPFS with symlinking. |
/hadoop/bigsql |
| BIGSQL_PORT | Db2 Fast Communication Manager (FCM) port number. | 28051 |
| DB2_PORT | Db2 connection port number. | 32051 |
| DB2_SSL_PORT | Port that Db2 uses to communicate with using SSL | 32052 |
| MLN_COUNT | Number of Multiple Logical Nodes (MLNs). Before or after you install Db2 Big SQL, you can specify a value greater than one to increase the number of
Db2 Big SQL worker node partitions.
Note: To change the number
of worker node partitions, you must use the Db2 Big SQL cluster administration utility.
|
1 |
| BIGSQL_HA_PORT | Port number used by Db2 Big SQL High Availability (HA). | 20008 |
| BIGSQL_HA_DB_BACKUP_DIR | Db2 Big SQL database backup directory when you are adding a second Db2 Big SQL Head. | /var/ibm/bigsql |
| CM_ADMIN_USER | Cloudera Manager admin user name. | admin Note: If you use a non-default Cloudera Manager user, the user must have
Cluster Administrator or Full Administrator privileges.
|
| CM_HOST | Cloudera Manager server host name. | |
| CM_PROTOCOL | Protocol (http or https) to access Cloudera Manager. | http |
| CM_PORT | Cloudera Manager port number. | 7180 |
| CM_SSL_CA_CERTIFICATE_PATH | When SSL is enabled for Cloudera Manager, the location of the truststore file. Note: If
Cloudera Manager is configured with its default auto TLS, set the value to
/var/lib/cloudera-scm-agent/agent-cert/cm-auto-host_cert_chain.pem.
|
|
| CM_BASE_CLUSTER | When there is more than one cluster defined in Cloudera Manager, the name of the cluster
where the HDFS, HIVE, Ranger, HBase, and Zookeeper services are installed. This parameter is not required if there is only one cluster defined in Cloudera Manager. |
|
| CM_COMPUTE_CLUSTER | When there is more than one cluster defined in Cloudera Manager, the name of the cluster
where to install Db2 Big SQL. If a value is not specified for this parameter, Db2 Big SQL is installed on the cluster that is specified in the CM_BASE_CLUSTER parameter. This parameter is not required if there is only one cluster defined in Cloudera Manager. |
|
| BIGSQL_MEM_PERCENT | Percentage of system memory that is assigned to Db2 Big SQL. | 25 |
| BIGSQL_SCHEDULER_ADMIN_PORT | Db2 Big SQL scheduler admin port number. | 7053 |
| BIGSQL_SCHEDULER_SERVICE_PORT | Db2 Big SQL scheduler service port number. | 7054 |
| BIGSQL_IMPERSONATION | Enables support for impersonation. | False |
| BIGSQL_PUBLIC_TABLE_ACCESS | When impersonation is enabled, grant select and IUD to public on CREATE HADOOP TABLE statement. | False |
| HBASE_ENABLED | Enable support for HBase. | False |
| RANGER_ENABLED | Enable support for Ranger. | False |
| ATLAS_ENABLED | Enable support for Atlas. | False |
| FC_ENABLED | False | |
| LDAP_HOST | LDAP server host name. | |
| LDAP_PORT | LDAP server port number. | |
| LDAP_BASE_DN | LDAP base DN. | |
| LDAP_BIND_DN | LDAP bind DN. | |
| LDAP_TLS_ENABLED | Enable TLS on the LDAP server. | |
| LDAP_ROOT_CA | LDAP root CA. |
bigsql-config
-enableRanger. To enable HBase, use the command bigsql-config
-enableHBase. To enable impersonation, use the command bigsql-config
-enableImpersonation.The log file for the utility is bigsql-config.log, and it is located in the /tmp directory.
Syntax
bigsql-config [options]The following [options] can be included when the utility is run by a superuser:
- -enableHBase
- Enable support for HBase.
- -disableHBase
- Disable support for HBase.
- -enablePam
- Enable support for Pluggable Authentication Module (PAM).
- -disablePam
- Disable support for PAM.
- -enableRanger
- Enable support for Ranger.
- -disableRanger
- Disable support for Ranger.
The following [options] can be included when the utility is run by the bigsql user:
- -display [all]
- Display configurable parameters or all parameters.
- -get key
- Display the current value of a configuration key.
- -set "key=value" [force]
- Update the configuration key to the value.
- -reset
- Reset to default values. You can use this option only before you install Db2 Big SQL. After the Db2 Big SQL installation is complete, you can't reset to default values.
- -enableAtlas
- Enable support for Atlas.
- -disableAtlas
- Disable support for Atlas.
- -enableImpersonation
- Enable support for impersonation.
- -disableImpersonation
- Disable support for impersonation.
- -help
- Displays the help.
Examples
- Specify the FQDN of the Cloudera Manager
host.
./bigsql-config -set "CM_HOST=<server_name>.com" - Set multiple disks for the data by specifying a comma-separated list of mounted database
directories:
./bigsql-config -set "DATA_DIRECTORIES=/disk_1/hadoop/bigsql,/disk_2/hadoop/bigsql,/disk_3/hadoop /bigsql" - Enable support for Ranger:
./bigsql-config -enableRanger