Saturday Jan 19, 2013

Journey to Database Cloud

Understanding the benefits of a database cloud usually leads to the question “How do I get there?” As the question itself implies, making the transition from a complex, legacy environment to a database cloud is a journey. Like any journey, clear goals and a plan to achieve them are the keys to success. Oracle’s “Journey to Database Cloud” is a maturity model that guides this process.

Each step of this journey delivers specific benefits. Knowing what you want to accomplish will help you identify the phase you need to implement. The key benefits and characteristics of each phase are:

Standardization: Simplify to reduce operational costs and business risk

Standardized deployments limit the number of environments and the processes that manage them to the smallest possible set of options. Hardware and software infrastructure are deployed in modular “building blocks.” Database versions are limited. Because each environment is simplified, each is easier to manage and maintain, which lowers operational costs.

A service catalog defines the deployment options for building blocks, and for the database configurations which end users may choose from. Because a small set of standard deployment patterns are followed, new deployments are easier to implement and can be activated with less risk.

Standardized components can be consolidated effectively since they can share a common infrastructure. And the higher degree of standardization that is applied, the higher degree of consolidation that can be achieved. Keep that in mind when evaluating the proposal from some vendors that transforming a datacenter into a private cloud is a simple matter of shuffling software stacks into virtual machines. This “quick fix” approach sounds attractive, but like most “quick fixes” it does not address the underlying problem of datacenter complexity. On the contrary, the added complexity of this approach results in lower standardization.

Consolidation: Efficiency reduces datacenter footprint - both hardware and software

In a traditional environment, servers are generally underutilized. Consolidating workloads onto shared infrastructure allows higher utilization and therefore a reduction in server footprint. This lowers both capital  and operational costs: lower power consumption and lower IT management expense since there are fewer physical environments to operate and manage.  Software environments are also decreased, meaning there are fewer software elements to purchase, monitor and maintain.

Several features and options of the Oracle database enable the consolidation of multiple databases onto shared hardware and software infrastructures. Features such as Database Resource Manager and Instance Caging facilitate sharing compute resources. Products such as Oracle Audit Vault and Database Firewall enable enterprise-grade security in consolidated environments.

That brings up another problem with the “build your cloud by putting everything in virtual machines” idea: virtual machines consume footprint and introduce performance overhead. They also require special skills and tools.  You can avoid those issues by consolidating directly onto the operating platform and achieve higher densities and better performance.

Service Delivery: Automate to increase business agility

Once organizations have successfully standardized their deployments and implemented their consolidation strategy, the next opportunity is to enable service delivery. By replacing manual processes with automated and dynamic capabilities, the environment responds quickly to changing workload conditions and requirements, which translates to faster operations and better agility. In the context of private database clouds, this means delivering Database as a Service (DBaaS).

In the standardization phase we defined a service catalog to describe the deployments to choose from. One focus of service delivery is to provide those choices via self-service, with as little manual attention from IT staff as possible. Self-service for end users allows them to choose from a menu of service options to create their own database environments online. This frees up IT for higher value initiatives.

Automated, dynamic management of resources is another key characteristic of a service delivery environment. In a consolidated environment managed with manual processes, adjusting a database’s footprint or resource allocation requires human intervention. Even noticing the need to make an adjustment requires human attention. By contrast, a service delivery environment uses tools to monitor and dynamically adjust resource allocations, without human intervention and without impact to running workloads. Features such as Oracle Quality of Service Management apply policies to monitor and manage workloads automatically and dynamically.

Enterprise Cloud: Unify services for location independence and unlimited capacity

One limitation of the service delivery environment is that workloads are bound to specific servers in a cloud pool. And while resource allocation within the pool is dynamic, pools are fixed in size (each one is typically one of the “building blocks” defined in the service catalog). This means that if a workload outgrows its pool, manual intervention will be needed to add compute resources to the pool, or to move the workload to a larger pool. Or, if the pool goes offline, there will be a service interruption while the service is reinstated on a different pool.

In an Enterprise Cloud, pools are dynamic and may grow or shrink as workloads dictate. Workloads are not bound to any specific pool, so if changes in workload patterns indicate that moving a given workload to a different pool is the best choice, the workload will be moved there without service interruption. Some workloads may even be distributed across geographically separated pools.

The ability for pools to grow and shrink dynamically and for workloads to migrate among pools of a unified cloud allows workloads access to virtually unlimited capacity. Since pools can be geographically separated, planned and unplanned outages of entire sites will not impact service availability.

The unified cloud is analogous to a utility service such as the electricity grid. Each user sees unlimited capacity and uninterrupted availability, and pays for resources as they are consumed.

Choice and Flexibility are Essential

You may choose to make a large change to reach your end goal, or you may choose to make incremental changes. For example, some customers have chosen to work with a standardized environment for some time before starting their consolidation efforts. Other customers made the move immediately to consolidation because they were keen on saving costs and floor space as aggressively as possible.

You might treat different workloads and environments differently. For example, one group within an organization may be technically and culturally ready to move their databases into a service delivery environment. Another group may be ready to standardize and consolidate, but not ready to implement service delivery. In this example, if the “consolidate only” group sees the tangible benefits the service delivery group enjoys, they’ll probably decide to make that step too.

You’re probably on the journey already…

While you were reading about the phases of the journey, you probably noticed several guidelines that you implement to varying degrees today. Standardization, for example, is a well-established approach that did not spring into existence when the industry starting talking about consolidation and service delivery. For standardization, what’s new in the context of the journey is recognizing the importance of this step and how the choices made here will have downstream impacts.

Wherever you are starting from and wherever you want to go in the journey to database cloud, Oracle has the products and services to get you there. And we have years of collaboration with customers in all areas of enterprise, government and education who have made this journey successfully, and continue to evolve their solutions with our cloud-enabling portfolio. We look forward to being a partner and mentor in your journey.


The Database Cloud Architecture Team at Oracle develops and documents best practices for designing and delivering database consolidation and database-as-a-service projects for on-premises database clouds.


  • Oracle
« January 2013 »