Distributed Database New Thinking: Cross-Cloud Platform

Distributed Database New Thinking: Cross-Cloud Platform

This is a simple solution that has a long history: using a distributed architecture in the database to quickly return request data. This solution runs database query operations on multiple servers at the same time, and then summarizes the returned results from hundreds or even thousands of servers in the cluster for delivery.

Even if it is not new, why has this idea gained new attention recently? This is because it becomes the core mechanism behind MapReduce, and MapReduce is the parallel processing mode adopted by Hadoop in big data analysis. These distributed workloads have been widely used in recent years. They are usually used together with even Server clusters-that is, they need to run on a large number of identical server devices. This uniformity requirement limits users to a single server cluster or a single cloud environment, which means we need to configure a resource type and cost scheme for it, and no other options.

HoweverIn a multi-cloud solution, the data processing workload runs on cloud services that can meet the actual needs of the workload.At present, the active exploration for multi-cloud architecture brings users the ability to place their workloads on public or private cloud services, so as to provide the shortest condition for their own load needs. This also allows everyone to run their workloads on the most cost-effective cloud services.

Process Data instances across multiple cloud platforms

For example, when the processing flow appears as a query, the client that starts this database query may reside in the running environment of the hosting service provider. However, it may direct requests to multiple server instances on the Amazon Web Services public cloud service. It can also manage transactional databases that reside in the Microsoft Azure cloud environment. In addition, it can save the data request results in the local OpenStack private cloud. Flexible and diverse. I believe everyone understands this feature.

The benefits of this solution are obvious: You can mix and match cloud services and workloads to improve performance while reducing the cost of use. In fact, we can migrate the workload between different cloud environments as needed.

Today, there are already a large number of database processing mechanisms that rely on cloud computing services-it is not cheap to use.Migrating workloads between different cloud services brings huge strength to users who manage large distributed databases., Allowing them to select only the vendors that provide the best and most cost-effective services-or those that best meet their database processing needs.

Of course, this type of transaction processing method is quite complex, and it is bound to put forward requirements for management and automation. Cloud management platform tools should play a role in this regard, because they can provide the capabilities of the actual processing process described earlier.

Keep a clear mind-this new idea is never as tough as it sounds, and having more choices is never a bad thing.

Http://www.infoworld.com/d/cloud-computing/the-right-cloud-the-job-multicloud-database-processing-here-247610

Original article title: The right cloud for the job: Multicloud database processing is here