Databases in the cloud, now on IBM Power Systems

Have you ever hoped for single-click database (DB) instance creation and configuration? Do you, like me, feel that cloud computing won’t be complete without database as a service (DBaaS)? How about automated provisioning of Oracle on IBM AIX, on-premises both with a standalone instance and Oracle RAC? Your dream can now come true.

Cloud computing isn’t a buzz word anymore; it’s more of a norm nowadays, with more organizations adopting it day in and day out. For private and public clouds, however, orchestration of enterprise-level databases into an integral cloud solution continues to be a challenge.

IBM Power Systems can help, incorporating both Oracle and Open Source Databases to provide a full as-a-service solution. Solutions include Oracle DB on IBM AIX as well as MongoDB, MariaDB, EnterpriseDB (EDB) and Cassandra DB with Linux on IBM Power Systems.

The challenge

Most cloud solutions are based on capturing software images to rapidly clone them into new copies that users can exploit. This task is not as simple with databases though, especially complex and deeply involved databases such as Oracle and EDB. It requires careful planning that incorporates additional installation processes that are normally required by the database when cloning an instance.

With Oracle, for example, there’s host-specific configuration data embedded inside the installed DB, and when captured and deployed, that configuration data wouldn’t be correct for the new instance. Also, with Oracle Automated Storage Management (ASM), the size of the database storage space in the image may not address all deployment requirements. Furthermore, when it comes to Oracle RAC, additional complexity is introduced for the cluster synchronization and homogeneity.

For Open Source Databases, there are a few challenges: difficulties can surface when rapid deployment is needed as well as around keeping up with the rapid versioning updates these databases adopt. Covering a wide range of both relational and NoSQL types could be demanding. Also, new database user IDs in the clone may be different from those of the image.

The answer

You can use IBM Power Systems virtualization and orchestration solutions to address these challenges. The image can be a generic VM with all of the configurations done externally and used by the installer upon cloning. Those configurations are set outside the image and are modifiable prior to cloning, to be honored during the cloning process. They dictate the version of the database to be installed, how much storage is needed, along with the rest of the necessary database instance customizations. So, ultimately, you end up with one solution that fits all (types, versions and sizes).

For Oracle, you can have a solution that handles the traditional deployment of 11g and 12c as well as the containerized version 18 of Oracle atop AIX. You can also use this technique for an automated Oracle RAC deployment.

For clients interested in exploring EDB as an alternative enterprise-level relational database management system (RDBMS), it’s quite simple and streamlined to get either version 9.6 or 10.2 of EDB. The solution currently also includes MongoDB, CassandraDB and MariaDB. Furthermore, adding a new Open Source Database to the solution package is straightforward.

The bottom line: Rapid database provisioning

IBM Systems Lab Services consultants have experience deploying “as a service” cloud database solutions both for Oracle on AIX as well as MongoDB and MariaDB on Linux on Power Systems. Our clients can leverage a streamlined process for rapid database provisioning as well as resource reutilization. The time for provisioning a new standalone Oracle ASM database instance can be reduced from hours or days to just minutes.

Are you ready to adopt?

If you’re ready to adopt a cloud database solution on IBM Power Systems, contact us today and it will be our pleasure to help you take your database provisioning to the next level.