Enterprises need to have a modulated strategy on workload management, CIO News, ET CIO

By Shailesh Kumar Davey

Today, digital is about automating processes both internally and externally and cloud has been the prime enabler for the same. The advent of newer technologies like AI / ML / blockchain, need for secure application access from anywhere, role-based access control of data, demand for self-service by employees, customers and partners, and requirement to tap into computing / storage power on a needs basis are some of the many factors that are driving organizations to shift towards the cloud.

Enterprises need to start taking gradual steps to address these aspects so that the organization does not end up becoming a laggard from both a technology and market perspective. For organizations the important factors that impact the decision-making pertaining to workload management are regulatory requirements, cost, possible future value, convenience, and talent availability.

Enterprises need to have a modulated strategy on workload management that is suitable for them and their business environment.

In terms of regulatory requirements, data sovereignty plays an essential role in the private or public cloud movement.

Organizations not on public cloud are instead leveraging private cloud. Though organizations have been using on-prem data center to manage apps and infrastructure, private cloud, with the usage of virtual machines and containers, software-defined infrastructure (network, storage, and compute) and administrative console, self-service control to DevOps personnel, etc is transforming the way organizations manage their compute infra.

For some of the applications, an easy way would be to retain the previous generation applications in a private data center and jump ship to a new SaaS-based application. The previous data could be imported into the newer SaaS application or could be referred to manually or via APIs when needed. Contrary to popular belief, a large category of applications surprisingly fall into this category.

For other applications that are already in a very homogeneous environment, suggests Davey, lifting them out of private data centers and shifting to co-location / cloud providers or public cloud could be a good learning and a low-risk experiment. Thirdly, modernizing in place and running as a private cloud could also be an option for some applications, especially if the regulatory environment demands so.

The most difficult option would be to refactor applications and turn them cloud-native.

Hardware refresh is not the only reason to switch to the cloud

The driving forces for organizations to migrate to the cloud includes cost, possible future value, convenience and talent availability.

In many cases, however, a hardware refresh cycle is definitely a key trigger as questions about migration come up at that time. The other factors could be the need for newer application functionalities, newer business models, and BI / analytics.

Brick and mortar retail stores may suddenly want to try an omnichannel selling model. A manufacturing enterprise may want to make its interaction with suppliers more digital, to increase business velocity. A financial company may want to adopt the newer fin-tech APIs in their application. Such application needs and market dynamics may force enterprises to reconsider their tech stack and adopt a cloud approach.

Must have cloud-based priorities for the future

In cloud-based enterprises, some of the previous organizational boundaries become less rigid and the need for DevOps and DevSecOps engineers arises. This calls for a cultural change in the IT department.

From a technology perspective, 5G, IoT, AI / ML workloads will call for a variety of deployments like ‘near-to-customer’ data centers, intermediate data centers and remote data centers. Accordingly, workloads will have to migrate between these to add maximum value to the customer. Additionally, technical requirements like latency, bandwidth, volume of data, etc., and security and privacy demands will play a major role in how these workloads are deployed.

The author is Co-founder and VP of General Management- Engineering, Zoho Corp.


Leave a Comment