Customizing Cloud Management to Suit the Workload

Ariel Maislos, CEO, Stratoscale

Ariel Maislos, CEO, Stratoscale

Now more than ever, organizations that transfer their resources to the cloud are looking for specific features. As cloud offerings mature and new business models evolve, requirements have become more sophisticated and precise. This has led to increased software specialization, such as the emergence of various PaaS offerings and “boutique cloud providers” that target specific industry segments and verticals.

Certain workloads are best suited for a specific cloud environment. Organizations that try the public cloud usually begin by migrating their less-critical workloads over and then gradually transferring the remaining workloads. Workloads that are best suited for the public cloud fall into two categories: batch workloads are easily automated and take advantage of the public cloud’s elasticity, while big data analytics benefit from the power and scalability of the public cloud, along with its burst capabilities.

Development and testing (DevTest) teams are also prime candidates for the public cloud because they are required to respond to cases quickly and in real-time. These cases can be removed when they are no longer required, which conserves resources and lowers costs.

In contrast, many organizations prefer to keep sensitive data secure on their own private cloud. IT teams may decide to place critical workloads into a private cloud to maintain control, higher availability and resiliency. Historically, workloads located on private clouds were vertically-scalable with long-term cases; however, as the private cloud matures technologically, it is able to support a wide variety of workloads. The private cloud is most effective when interdependent workloads require minimal latency for data access because workloads can be located in close proximity to each other.

​  As the private cloud matures technologically, it is able to support a wide variety of workloads. 

Some workloads are better off located in legacy data centers, such as those that would require costly and time-consuming rewrites to migrate them to cloud environments. Organizations that want to benefit from the cloud, but have many legacy applications, often find that a hybrid solution allows them to keep their workloads on the infrastructure that best suits each individual workload.

As more organizations are moving some or all of their workloads to a cloud environment, the challenge becomes how to best manage these workloads from an end-to-end perspective. Arthur Cole of IT Business Edge notes that, “cloud management platforms will have to be very ‘workload aware’ in order to deliver the performance, availability, optimization, and IT operations analytics that organizations demand in order to track cloud service level agreements (SLAs) and to solve end-user issues.”

While businesses have adopted the cloud for its agility, scalability, quick response time, and self-service features, moving higher-level and critical workloads to the cloud requires complex solutions. Synchronization between the various elements across the application-both physical and virtual layers-will lead to the creation of more specialized offerings by public and private cloud providers. This may lead to multi-cloud and hybrid solutions-as different clouds are required for various workloads.

Read Also

The "Black Box Paradox" in Big Data Analytics and Data-Driven Modeling

Daniel Lingenfelter, Staff Engineer, Seagate Technology

A Look into the Disrupting Industry of Cyber Space

Jerrod Chong, VP of Solutions, Yubico

The Storm Behind the Cloud: Ushering In the Next Era of Innovation

Merijn te Booij, Chief Marketing Officer, Genesys

Cloud: Enhancing Truly Unified Communications for Businesses

Ken Bisnoff, SVP of Strategic Opportunities, TelePacific