Key points

• Rapidly deployed compute capacity to meet changing demand
• Simplified aggregated bandwidth billing with Zenlayer load balancing
• Deployed thousands instances with clear cost breakdowns and responsive technical support

 

About the customer

Industry: Enterprise web data services
Needs: Flexible compute, 100Gbps traffic target, support for a large IP footprint
Zenlayer services used: Zenlayer Elastic Compute (ZEC), load balancer, BYOIP, and Premium BGP bandwidth

The customer is an established enterprise web data aggregator serving enterprise and SaaS clients. They help customers collect public web data at scale, moving petabytes of traffic each month. Well-managed IP space, reliable network performance, and close control over infrastructure costs are all paramount to scaling their business.

 

The challenge: supporting a large AI client within a tight budget

The web data aggregator initially came to Zenlayer for a bare metal quote to support a large new AI client as their project needed infrastructure sized for 100Gbps with compute instances that could be spun up and down on the fly.

However, since their client’s ramp was still uncertain, their final compute footprint was difficult to predict.

Their CPU estimates came from previous deployments, so they needed help working through the requirements before their new deployment. They also wanted the flexibility to add capacity quickly and expand into more regions like the U.S. and Europe.

Beyond compute, they planned to bring their own IP address (BYOIP) capabilities and wanted clarity on how traffic routing, bandwidth, and load balancing would impact their bill.

 

Matching compute and billing to the workload

We worked with the web data aggregator on sizing and traffic expectations as well as daily deployment needs. ZEC gave them the flexibility to provision and release instances on demand and helped them seamlessly add capacity across global locations on one platform.

Most of their existing infrastructure ran on bare metal. With ZEC, they can now manage compute and network resources through zenConsole and use built-in services like managed load balancing, simplifying their day-to-day operations.

At a 100Gbps target, they needed to ensure predictable ongoing bandwidth costs. We reviewed the available options, including flat-rate bandwidth for unmetered use.

We tailored a simplified, predictable billing structure and provided a cost sheet for a live instance, broken down by compute, load balancing, bandwidth, and other charges, giving them a clear basis for planning costs and comparing providers.

Following cost analysis, they are now migrating additional workloads to Zenlayer.

 

Simplifying large IP deployments

Their IPv4 addresses were mostly in /24 blocks. Before moving any prefixes, they needed to know what onboarding would entail and whether our load balancer could support their planned address count.

With capacity confirmed, we walked them through ZEC’s BYOIP and BYOASN requirements. Our network operations team also clarified the onboarding process, with most steps available through zenConsole self-provisioning when needed.

With traffic spread across such a large number of instances, they needed a simpler way to manage bandwidth and avoid unexpected overage charges. We configured their ZEC instances behind our managed load balancer, which acted as a central aggregation point to simplify their traffic flow and bandwidth billing.

 

Supporting the deployment as it grows

The web data aggregator needed capacity ready as their deployment grew, so we stayed in regular contact about upcoming deployments and regional plans and expanded the available instance configurations to support their growth. They have now deployed thousands of ZEC instances across our platform.

Given the scale of their deployment, our VP of Product stayed involved throughout the process and helped guide a redesign of their failover workflow to better balance traffic across multiple ASNs.

We continue to support them after their deployment. When an upstream carrier issue arose, we identified it before they did and handled it directly. The customer expressed appreciation for our responsiveness, especially outside normal business hours.

They now have a repeatable deployment template they can reuse as they expand into new markets.

 

Looking ahead

Expansion into US and Europe is underway with Asia and LATAM planned next. The enterprise web data aggregator now has a workable path toward their 100Gbps target, and we’ll continue helping them support additional large AI data customers.

 

Simplify global AI deployment with Zenlayer AI Gateway

Our Elastic Compute, premium BGP bandwidth, and hands-on infrastructure guidance helped this enterprise web data aggregator deploy thousands of instances while keeping capacity flexible and costs predictable. We can help you do the same.

If you’re looking to scale data-intensive workloads, manage bandwidth costs, or expand into new markets, talk to a Zenlayer solution expert today.

For instant provisioning, check out zenConsole, our self-service platform that lets you deploy compute around the world in minutes.