Over the last few years my team has helped a number of key and strategic customers with how organisations with widely deploy stand-alone Oracle database environments find a more cost effective and efficient solution for the delivery of database services.
More and more we see each corporate initiative or project acquires its own hardware & software and establishes separate hosting & support arrangements. Consequently, there has been (and will continue to be) a proliferation of servers and software that all fundamentally do the same thing - provide Oracle database services for applications to consume.
A more effective and cheaper way of providing Database Services is to "pool" several independent database servers together and produce an Oracle as a Service (OaaS) grid. In this architecture, applications would connect to a single shareable instance of Oracle as they would any other Oracle instance. The business applications would be isolated from each other and explicit portions of Oracle processing power allocated to each one - a high-use application would be granted a large portion of processing power, a low-use application would be granted less. This provides Oracle customers with a number of realisable benefits:
- Higher availability; as all participating applications immediately benefit from automatic failover infrastructure.
- Cost savings; significant capacity for savings on hardware through better hardware utilisation and the use of cheaper commodity servers.
- Better service through centralised management; organisations can benefit from higher quality by leveraging a central team of database administrators.
- Reduced risk; the entire environment is managed, support is centralised and process, procedure and skills are standardised.
OaaS Design Pattern
The OaaS design pattern has been applied at a number of strategic customers that have delivered significant business benefit and is typically deployed as follows:
- Construction of one, or more, clusters consisting of standardised commodity hardware components that hosts these shareable Oracle instance
- Establishment of procedures for project & application support teams to engage with the Oracle shared services team and have their database applications hosted in this environment
- Definition of a cost, funding & operational model for the shared Oracle environment and the organisations IT operations team.
Reasons to consider Consolidation
In the current economic climate, keeping recurring operational costs down is imperative. Consolidation will assist the reduction of total cost of ownership (TCO) in several ways, including envisaged savings in hardware, software and hosting, as follows:
- Reduced administration. By standardising and reducing the number of servers, businesses reduce the complexity of the infrastructure they must administer. Fewer support staff can therefore manage the same service demands. This standardisation also facilitates the provision of 24/7 support using worldwide support resources.
- Reduced operational costs. Increase in service capacity and growth are achieved with better utilisation of resources. Typically, fewer servers are required, resulting in hardware and power savings.
- Reduced data centre costs. Site space formerly used for IT services can be returned to the business, and existing facilities can be used more efficiently.
- Reduced data centre hosting costs. Typically, hosting contracts are based on infrastructure under management - the less the infrastructure under management, the less the cost.
- Reduced revenue loss through higher uptime/availability. Consolidation reduces the cost of implementing high-availability solutions, which reduces revenue losses due to downtime.
- Improved service management. By standardising and reducing the complexity of service infrastructure, organisations facilitate more effective service management processes, tools, and automated system administration.
- Simplified contingency planning solutions. A simpler service infrastructure means that more services can be restored when a site failure occurs.
- Improved quality of service (QoS). Consolidation will facilitate improvement in application up-time and will deliver higher systems performance.
- Increased reliability and availability. Server consolidation makes it more economic to provide high-availability configurations and dedicated support staff. Organisations also benefit from implementing better storage management and service continuity solutions.
- Improved performance. Standardised systems deliver more predictable performance, and proprietary technologies such as data compression can improve query response times.
- Improved infrastructure agility. In uncertain times, flexibility and the ability to respond quickly to changing needs of the business can help to maintain the competitive edge. Consolidation results in a more standardised, centralised and dynamic infrastructure, which makes it possible for systems to be more responsive to change and to quickly adapt to business needs.
- Improved consistency. Consolidation improves interoperability, and makes systems more productive and easier to manage:
- Better integration. A consolidated platform provides for easier, and cheaper, systems integration, which improves data consistency and reduces the complexity of tasks such as extract, transform, and load (ETL) operations.
- Centralised management. Because consolidation facilitates centralised management, it becomes easier to implement standard policies across your systems.
- Reduced carbon footprint. With greater emphasis on sustainability, many organisations are striving to reduce the environmental impact that their activities have. Consolidation enables organisations to reduce their energy consumption by using less hardware as well as by using that hardware more efficiently.
- Improved resource utilisation. Most applications that utilise Oracle elect to provision a server/s with capacity that will exceed their maximum peak load. It is a global phenomenon that many database servers run at low utilisation rates. In a consolidated deployment model, idle capacity across this infrastructure can be reclaimed by reducing the total number of required servers while not affecting application performance or availability (although the extent of the envisaged savings need confirmation within each particular organisations environment).
OaaS – The way forward
Traditionally the type of application workload for the database was a key consideration in designing and configuring the environment as allocations for, and use of, resources can be very different depending on whether the workload type is Online Transaction Processing (OLTP) or Decision Support System (DSS).
In October 2008, Oracle unveiled the Oracle Exadata family of products which simplifies the deployment of an OaaS platform. The Exadata family of appliances are geared towards providing high-performance, ready to scale database processing capabilities specifically mixed applications workload (DSS or OLTP) and simplifies the first step in the Design Pattern, by providing a pre-configured standardised platform.