Preparing for outages key when maintaining cloud database systems
When a business establishes a database within its data center infrastructure, it usually will put systems into place to ensure that solutions is available at all times. Strategies to accomplish this include having redundant power supplies in place, having the database duplicated in a secondary server in case the primary hardware fails and having the entire setup backed up in a remote facility in case that data center experiences an outage. Network considerations also are taken into account and the entire backup process is complex and multi-tiered.
However, many companies tend to act as if, once they put their databases into the cloud, they no longer have to worry about backing their systems up. Sometimes this is the case, but organizations do need to be prepared for an outage on the part of the cloud provider. In many cases, the cloud service is far more reliable than most corporate IT systems, but outages happen and are often unpredictable, and companies need to be ready whether they are in the cloud or not. A recent PCWorld report explained that cloud systems, whether databases or application hosting solutions, can be managed in such a way that organizations can establish backup solutions and redundancy so they can ensure availability.
One strategy companies can use is making sure their cloud systems are maintained in separate regions, the news source said. Geographic diversity is key when trying to ensure cloud availability because natural disasters, power outages and similar issues, which are local in nature, tend to be some of the primary causes of unplanned downtime. Having a secondary cloud setup in a remote location from the primary platform can make it much easier to ensure availability.
Drue Reeves, a Gartner cloud analyst, told the news source that organizations working for a more reliable cloud may also want to consider establishing accounts with multiple vendors. This provides an inherent level of availability as organizations can simply switch between systems if one of them experiences an outage. For example, they can devote many of their resources to a primary cloud vendor to meet performance, scalability and flexibility requirements, and use a low-budget option that can simply maintain availability as the backup. However, Reeves warned that some cloud vendors share data center resources, as a result, businesses need to verify that they do not use separate cloud providers, only to find out that the two vendors use the same data center.
Adjusting service level agreements to focus on availability can also help. The report explained that negotiating an SLA that is focused on availability can provide the foundation necessary to help businesses ensure that their systems are available at all times, or at least as close constantly present as possible.
The key to understanding reliability in the cloud is to split responsibilities appropriately. As organizations work to move their proprietary and open source database systems into the cloud, they also have to consider the measures they need to take to keep those solutions available. However, the efforts to ensure availability in the cloud are not all that different than those needed to do the same in premise-based systems. As a result, companies do not need to look at potential cloud failures as a reason not to implement the technology. The risk of an outage is prevalent in any IT system, and a cloud vendor can generally offer better availability than most companies because it is one of their core competencies.