The Key Decisions When Migrating Your Databases to the Cloud
How do databases and the cloud get along? This has been a recent topic of discussion with many businesses that are focusing on monetizing their data. If you’re thinking about taking your organization’s databases to the cloud, there are a couple of approaches you could take.
Databases can be deployed to most cloud providers using Infrastructure as a Service (IaaS) or Platform as a Service (PaaS). Software as a Service (SaaS) is another option, but it’s typically used for applications and rarely leveraged for database management systems.
Adopting the IaaS Model
IaaS is very similar to deploying a database in your existing data center with virtual machines. You must define or configure your network, storage, servers and virtualization in an IaaS model, but the process of acquiring physical hardware is already resolved. Like owning a physical data center, you request a server and install your operating system (OS) and database software as you would on any server.
A key to a successful IaaS implementation is the speed of delivery from your on-premise implementation. In most cases, you can create an identical environment to your on-site implementation and then migrate. There are drawbacks to this approach.
First, your choices for high-availability and disaster recovery are limited to more traditional methods. And second, depending on the database system, your choices may be limited or incompatible with methods used in the cloud-based infrastructure.
Leveraging the PaaS Model
PaaS is used to manage your database and OS. In many cases, these implementations are lighter-weight versions of the database software and only require some configuration. While PaaS lacks some functionality, it spares you from managing physical hardware. For instance, an on-premise data center requires your staff to purchase and build your physical infrastructure and private cloud that emulates PaaS services.
If the underlying infrastructure of your PaaS provider differs from your current database engine, you’ll need to adopt their model for database management. Though this shift will also require a change in your high-availability planning and configuration, it allows you to take advantage of your PaaS provider’s many data centers. You will be able to quickly shift resources from one data center to another without losing functionality. But remember, in order to take full advantage of these benefits, you may have to change how you manage and implement your database functionality or lose some capabilities altogether.
What Approach is Best for Your Business?
The ideal database management strategy depends on your organization’s needs and the critical level of your systems. Before migrating your database, you’ll need to assess:
- Your organizational aversion to risk during your migration
- The implications on how much of the traditional database engine an application is leveraging for functionality
- What changes will be required to accommodate your new database engine in the cloud
- What operational procedures and processes must change to optimize your cloud offering
As a Senior Enterprise Data Architect at Integress, I specialize in Database and BI Architecture, Development and Design, Data Quality and Integration, Process Design, Application Integration, Unix Shell Programming, and Statistical Modeling.
Email me to further discuss data migration opportunities.