Database Configuration

Configuring ODAS to use a remote database server is recommended for all ODAS clusters.

Configuration Settings

The following configuration settings are used to configure a remote database server:

• CATALOG_DB_ENGINE
• CATALOG_DB_URL
• CATALOG_DB_USER
• CATALOG_DB_PASSWORD
• CATALOG_DB_HMS_DB
• CATALOG_DB_SENTRY_DB
• CATALOG_DB_USERS_DB
• CATALOG_DB_OKERA_DB

Configuring the Database

There are two parts to configuring the database:

1. Configuring database server connectivity
2. Configuring what databases to use on the server

Configuring Database Connectivity

To configure the connection to the database, you will need to know the host and port of the database server, and the username and password for a user that has sufficient privileges.

For example, if your MySQL database is at db.example.com:3306 with a username/password of okera/Okera123!, the configuration will look like this:

CATALOG_DB_ENGINE: mysql
CATALOG_DB_URL: db.example.example:3306
CATALOG_DB_USER: okera
CATALOG_DB_PASSWORD: Okera123!

Configuring Database Names

On your MySQL database server, ODAS will use four databases to store different types of values, such as the HMS catalog, policy information, user storage, etc. These databases have the following default values:

• CATALOG_DB_HMS_DB: hms_db
• CATALOG_DB_SENTRY_DB: sentry_db
• CATALOG_DB_USERS_DB: cerebro_users
• CATALOG_DB_OKERA_DB: okera_db

These values can all be changed. For example, to modify them to have the prefix of example_, you would have the following in your configuration file:

CATALOG_DB_HMS_DB: example_hms_db
CATALOG_DB_SENTRY_DB: example_sentry_db
CATALOG_DB_USERS_DB: example_cerebro_users
CATALOG_DB_OKERA_DB: example_okera_db

Note

You can edit only some of these settings - it is not required to set all of them.

Required Privileges

The user account that is used to connect to the database server must have sufficient privileges to fully administer the above four databases. This is typically achieved by providing superuser credentials that has full privileges for the entire databases server, but can also be done by creating a user that has admin privileges on only those specific databases.

For example, to grant a user called myuser access to the example_... databases above in the MySQL shell:

> GRANT ALL PRIVILEGES ON example_hms_db.* TO 'myuser'@'%'
> GRANT ALL PRIVILEGES ON example_sentry_db.* TO 'myuser'@'%'
> GRANT ALL PRIVILEGES ON example_cerebro_users.* TO 'myuser'@'%'
> GRANT ALL PRIVILEGES ON example_okera_db.* TO 'myuser'@'%'

Verifying DB Connectivity

To verify that the ODAS pods can access the MySQL database server, you can launch a shell inside the Catalog pod and use the MySQL CLI to connect to the server.

1. Run kubectl get pods to get the list of current pods and find the one starting with cerebro-catalog-...:

   $ kubectl get pods
   NAME                                 READY   STATUS    RESTARTS   AGE
   canary-fdtmn                         1/1     Running   0          5d16h
   canary-qz56w                         1/1     Running   0          5d16h
   canary-vlknv                         1/1     Running   0          5d16h
   cdas-rest-server-799dbd7ff4-tfr57    1/1     Running   0          5d16h
   cerebro-catalog-7f58569c4d-gv46c     1/1     Running   0          5d16h
   cerebro-planner-69c977496-wbzc4      1/1     Running   0          5d16h
   cerebro-worker-8xzbw                 1/1     Running   0          5d16h
   cerebro-worker-fzsps                 1/1     Running   0          5d16h
   cerebro-worker-jspr2                 1/1     Running   0          5d16h
   presto-coordinator-f698dfd7d-lfrhf   1/1     Running   0          2d23h
   presto-worker-glzv2                  1/1     Running   0          2d23h
   presto-worker-xbxv9                  1/1     Running   0          2d23h
   presto-worker-xdg58                  1/1     Running   0          2d23h
   zookeeper-589f995984-fwlv8           1/1     Running   0          5d16h
   

2. Run kubectl exec -ti cerebro-catalog-... bash (replace the pod name with the actual value from the prior command) to launch a shell inside the pod:

   $ kubectl exec -ti cerebro-catalog-7f58569c4d-gv46c bash
   root@cerebro-catalog-7f58569c4d-gv46c:/#
   

3. Run the mysql CLI to connect to your database server, e.g. if the MySQL database is at db.example.com:3306 with a username/password of okera/Okera123!:

   root@cerebro-catalog-7f58569c4d-gv46c:/# mysql -h db.example.com -P 3306 -u okera -p
   Enter password:
   Welcome to the MySQL monitor.  Commands end with ; or \g.
   Your MySQL connection id is 38422
   Server version: 5.6.39.0 MySQL Community Server (GPL)
   
   Copyright (c) 2000, 2018, Oracle and/or its affiliates. All rights reserved.
   
   Oracle is a registered trademark of Oracle Corporation and/or its
   affiliates. Other names may be trademarks of their respective
   owners.
   
   Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
   
   mysql>
   

If the connection hangs, it may mean that network access is blocked between your ODAS nodes and your database server (typically a network configuration issue, e.g. a security group on AWS or Database Firewall on Azure).