Skip to main content
Cloud

What is vertical scaling?

And how to know when to scale up.

• 4 min read

TOPICS: Cloud / IT Infrastructure & Operations / Cloud Operations

We’ve all been there. Your website is hit by a spike in traffic and suddenly, you’re faced with one burning question: How will you increase your server capacity?

For IT pros seeking a temporary fix, vertical scaling may be the answer. Vertical scaling (a.k.a., “scaling up”) is when more resources (RAM, CPU, etc.) are added to a system or server to increase its capacity, as opposed to upping capacity through additional machines. (The latter is known as horizontal scaling.)

“The way I like to think about it is just maximizing a single host or server or instance to get the most out of that one thing instead of creating a more complicated environment with multiple redundant hosts,” Ben Bernstein, manager of cybersecurity advisors at Huntress, told IT Brew.

What’s it good for? According to Osmar Olivo, senior director of product management and core database at MongoDB, a company may choose to scale up because it is often quicker and simpler than horizontal scaling.

“At the end of the day, particularly with the hyperscalers, they make it as simple as just clicking a button and saying, ‘Please move my workload, please move my application, and please move my database to a machine that is twice as large as this one,’” Olivo said. “And you just press a button, your build changes, and suddenly your application, your services, have twice as many resources at your disposal.”

Bernstein added that there are fewer overhead costs associated with vertical scaling, as well as reduced licensing costs, given how software licenses tend to be charged on a per-server or per-node basis.

It also has its cons. Similar to vacationers at the airport, scaling up comes with its own baggage. For one, vertical scaling creates a single point of failure; if one server goes down, the entire system comes crashing down with it, leading to downtime.

“Whether you’re technical or not, that sounds bad,” Bernstein said. Vertical scaling often requires companies to plan for downtime because servers need to be taken offline for upgrades.

Olivo added that vertical scaling can be costly, especially when a company approaches the maximum capacity a machine or system can handle.

“You can buy a machine that’s twice the size, and that’s going to not necessarily cost you twice as much,” Olivio said. “That might cost you 2.5x as much, or 3x as much. As you continue to progress higher and higher, that number just starts to get further and further away from linear.”

Top insights for IT pros

From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.

By subscribing, you accept our Terms & Privacy Policy.

Finally, vertical scaling has physical limitations as to how much a machine can be upgraded, according to Olivo.

“The limit was perceived to be very large, but now that we’re entering this agentic scale world where we’re having AI agents write twice as much, read 10 times as much, [and] process 10 times as much, I think it’s becoming more and more common that people start to run into the boundaries of just how big a machine can get,” he said.

Ways around? Are there ways to evade some of the shortcomings that come with vertical scaling? Alan Rencher, CTO at dental software platform Henry Schein One, told IT Brew that businesses in a crisis and trying to scale up likely have “no choice” but to swallow costs.

However, Rencher added, organizations could work around any single point of failure created by that scaling-up.

“You can do failovers and redirects and things, but there’s still a blip,” Rencher said. “You’ll be down and get error messages for a couple minutes, five minutes.”

And that downtime comes at a cost. “If you’re buying something online and you go to their site and you get, ‘Sorry, out of service,’ you automatically say, ‘What kind of janky company is this?’” Rencher said. “The reputational damage and the risk is way too great. That’s been my experience.”

When to use vertical scaling. When does it make sense to scale up? Olivo said it might be easier to rely on vertical scaling during times of crisis or an unexpected load traffic spike.

“Vertical scaling is by far the fastest way to get you out of a situation,” Olivo said. “But what happens if you’re already on the biggest box, and there’s just nowhere for you to go? In that world, you don’t have vertical scaling as an option. Horizontal scaling is your only option.”

Businesses with the wiggle room can also treat vertical scaling as an “emergency escape valve,” meaning organizations can scale up during times of large growth and scale back down when appropriate, he added.

“Are you really going to double, triple, quadruple your operating costs and stay at that level for an extended period of time? Probably not,” Olivo said. “And that’s something that you know we advocate. We see a lot [of] people scaling up and then scaling back down.”

About the author

Brianna Monsanto

Brianna Monsanto is a reporter for IT Brew who covers news about cybersecurity, cloud computing, and strategic IT decisions made at different companies.

From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.

By subscribing, you accept our Terms & Privacy Policy.