Rethinking Workloads for the Cloud Era

Workload is, well, a loaded word. It’s frequently and widely used by IT professionals, but it can mean different things to different people, especially in a world increasingly powered by a variety of cloud computing technologies. One thing’s for sure, all workloads and no play can be a daunting way of life.

“The word workload presses into my chest and fills me with dread,” said Steve McDowell, senior tech analyst at Moor Insights & Strategy. “Load takes up as much space as work, and gets the final say. Load implies a burden. Effort. Something to avoid.”

In computer science terms, a workload refers to a computer system’s ability to handle and process work. Viewing, editing or sharing a photo from a laptop requires the computer to process software instructions. Every time someone makes a search on Google, a workload is processed by data centers that find and present results.

Simply put, a workload is about “putting elements together to get data, finding out what something means or developing something,” said Judith Hurwitz, president and CEO of Hurwitz & Associates and author of Cloud Computing for Dummies. “It’s fundamental to computing.”

[Related story: Metadata Unlocks the Wonders of Data]

As more aspects in business and life become digitized, the number of computing workloads is skyrocketing. To handle the rising demand, McDowell said new innovations help “offload” work.

“As technologists, we can break workloads down into the processing piece, the storage piece, and the networking piece,” said McDowell. “It requires balancing those three things.”

For businesses and organizations relying more on data technologies, he said it’s critical to have IT infrastructure or leverage services that manage each of these pieces effectively and efficiently.

The real challenge is providing the ability to manage the disparate pieces as a whole. This is where IT organizations should look to vendors who deliver a management plane that spans storage, compute, and networking. This should work both within the datacenter, as well as across a hybrid-cloud environment.

McDowell said the skyrocketing rise of workloads is driving many to experiment and run their businesses on cloud computing, which can be rented or built on premises. It allows for scaling up or down compute, storage, and networking to meet changing needs.

Define Your Workloads

A workload is basically something running somewhere providing a service to the consumer. They come in many flavors and sizes. What's important is what they're doing and what services they are providing, according to Steven Poitras, principal solutions architect and author of The Nutanix Bible.

“It could be anything from a system running analysis on field samples on an oil rig to a transactional database handling order management for a business,” he said.

Some workloads can be static - always there, never off. Core services like enterprise resource planning (ERP), customer relationship management (CRM) and email systems are examples of static workloads. The demand and uptime requirements are known.

Poitras said there are also dynamic or ephemeral workloads, things like spinning up a PoC environment for a new project or performing batch processing at the end of the month.

“The importance of workloads will always vary based upon the context (e.g. who's being affected),” he said. “However, from a business perspective, anything that hits the top line is definitely a higher priority vs. items of convenience.”

Where to Run Workloads

Cloud is extremely attractive for projects that need to be spun up quickly, or projects that have a limited life, said McDowell.

“Moving them to the cloud delivers flexibility,” he said. “You can provision infrastructure, whether storage, compute, or networking, with the press of a button. Payment for public cloud services is based on usage, not capital investment.”

Workloads get highly distributed through cloud computing, said Hurwitz, which increases efficiency.

“It doesn’t just sit on a server somewhere and crank along and do what it has to do,” she said. “The workload may move into the platform or service that is best to execute at the right speed.”

But compliance and latency can complicate cloud-based workloads, said Hurwitz, requiring some data to remain in specific parts of the world.

“Some workloads can only live in a certain geographies due to compliance issues,” she explained. “Or if you need to accomplish something faster, that workload should execute its function very close to the source of the data because you don’t want latency.”

[Related story: 3 Laws Disrupting Data Management]

Not all workloads should run in the cloud, so Hurwitz said it’s important to balance business needs against costs.

Gone are the days where IT practitioners would manually layer storage, servers, and software together for each workload. Today we are in the midst of a revolution in intelligent software that can assist in both provisioning and adapting infrastructure to workloads as needed.

[Related: Make Infrastructure Invisible with Nutanix]

Where and how workloads and services are consumed can change, said Poitras.

“For example, we now can consume customer relations management data (CRM) or IT service management (ITSM) as services delivered by the vendor,” he said. “In this case the service stays the same, however, where the ‘workload’ will run is what changes.”

Workloads are always evolving. Some may come and go and others can move somewhere else, according to Poitras.

“It makes perfect sense to consume some workloads as a services and push the burden to the vendor,” he said. “However, for others, it may make the most sense to keep workloads running internally on-prem or hosted in the cloud, depending on the level of customization or concerns around data gravity and privacy.”

Poitras thinks about what needs to be consumed – like email – at the front-end. Then there’s the technology in the background required to deliver that service.

“To provide that service I may be running an email Exchange workload on premise or I directly consume a service like Office 365 or Gmail,” he said.

He said cloud brought things like infrastructure as a service (IaaS), platform as a service (PaaS) and software as a service (SaaS), which gave IT teams a lot of flexibility to directly consume services like ServiceNow and Workday. Or it [cloud] provided an environment in which to run these services.

For certain services, a workload may evolve.

“For example, a reporting service may have initially been a report running on an OLAP database that may have evolved into a Hadoop/ElasticSearch workload, which could have evolved to run on Amazon EMR,” said Poitras.

Knowing how and where to run workloads is more complex than ever, but cloud gives IT an opportunity to think differently. Increasingly, there’s more flexibility and choice, and software innovation keeps racing to simplify the complexity.

“Machines can manage the more burdensome efforts,” McDowell said. “That allows the rest of us IT folks to focus on more interesting problems.”

Stephanie Vozza is a contributing writer who specializes in business and productivity. She is a columnist for FastCompany.com [fastcompany.com], and her byline has appeared in Inc., Entrepreneur and Success magazines. Find her on Twitter @StephanieVozza.

© 2019 Nutanix, Inc. All rights reserved. For additional legal information, please go here.