Why companies should explore open-source solutions for an interoperable, multi-cloud strategy

By Siva Padisetty

A typical innovation play for technology companies will follow the same path:

• One company will develop innovative, proprietary technology;
• Based on the demonstrated success of the first company, multiple companies will follow with alternate proprietary technologies;
• Reacting to customers’ deep-rooted interest in not being locked in to one technology, new waves of companies will try to build abstraction layers on the proprietary technology;
• The initial company will likely fail because they cannot keep up with the underlying pace of innovation;
• As the technology matures and open-source technology catches up, the industry will move towards standardization; and
• Based on lessons learned from past failures, new companies will succeed with the standardization.

If you look around, you can find several examples of this evolution across different industries.

Think about Apple and their Lightning port. By using a proprietary connector, Apple was able to take control over the accessory ecosystem, gain higher profit margins from licensing fees, and protect their ability to design for specific form factors and functionalities. But, over time, Apple couldn’t hold onto this position. The consumer electronics industry adopted USB-C as a standard, consumers began demanding universal charging stations, and regulators in places like the European Union (EU) forced Apple to adopt the new standard instead of their proprietary technology.

We’re now seeing this pattern come to a head with cloud providers. Companies that don’t react quickly will be locked into a single cloud, losing the agility they need to continue growing and innovating. Those who can plan and execute a forward-thinking multi-cloud strategy will be better positioned for the future.

Companies should carefully explore and define their multi-cloud adoption strategy. This is often a multi-year journey, and I recommend companies take a crawl-walk-run approach to their transformation. This article captures some of the different ideas and approaches companies can take as they begin their multi-cloud adoption.

Aligning your strategy with business objectives

A multi-cloud adoption strategy won’t happen by accident. Making this change will require a company to unwind existing practices and reimagine new ways of building software.

When designing a multi-cloud approach, start by defining the business objective driving the decision. There are several factors that could be motivating the transition:

• Not getting locked into a single cloud provider and maintaining flexibility for future innovations
• Achieving negotiating power with a current cloud provider and driving down overall cloud costs
• Improving agility by leveraging open source in targeted areas of the tech stack

The many tools and technologies available in the cloud-native landscape can often lead to fragmentation and vendor lock-in. To counter that trend, companies should explore open-source initiatives.

One of the most important open-source cloud initiatives is the Cloud Native Computing Foundation (CNFC), which hosts several crucial components of the global technology infrastructure. CNCF is responsible for Kubernetes (for containers), OpenTelemetry (for observability), and Envoy (for service mesh), ensuring that each of these technologies remains open and accessible. As a result, companies can benefit from a standardized, interoperable ecosystem that supports multi-cloud strategies.

It’s important to note that companies shouldn’t try to achieve a 100 percent cloud-neutral design. This approach requires companies to align their technology with the lowest common denominator, resulting in increased complexity and decreased innovation. Companies should instead decide which part of their tech stack should be cloud-agnostic based on the availability of open-source alternatives and the relative cloud cost of each component.

When evaluating open-source alternatives, companies should consider both the migration cost and the ongoing operational cost. Choosing to replace the components that generate relatively high cloud bills enables companies to free up a bigger budget that they can control rather than leaving it up to the mercy of their cloud provider.

How New Relic approached a multi-cloud strategy

At New Relic, we went through our own journey to adopt a multi-cloud strategy.

New Relic can be broken down into a control plane and a data plane. The control plane is complex and contains the logic that configures how telemetry data is processed and visualized; the data plane handles the massive scale of processing and storing exabytes of data. The control plane accounts for approximately 10 percent of our overall cloud costs, while the data plane accounts for the remaining ~90 percent.

We developed a massively scalable data store called the New Relic database (NRDB); the NRDB is cloud agnostic and relies on a number of open-source technologies such as Kafka and Kubernetes. Because the NRDB is cloud agnostic, we can maintain low development and maintenance costs: The data plane comprises 20 percent of our development costs, as opposed to 80 percent for the control plane. For New Relic, the smart strategy was to pick cloud-agnostic technology for the data plane and cloud-specific technology dependencies for the control plane.

Like most companies, agility and innovation are key to our long-term success. In some cases, picking cloud-specific technology can help enable a company to maintain a competitive edge.

When we were developing New Relic AI, we chose to use OpenAI from Microsoft Azure. Artificial intelligence (AI) is a fast-evolving, disruptive technology; waiting for the tech to be standardized would have been a strategic mistake. Standardization will come over a longer time frame, so the key was to build an AI tech stack in a way that allows our organization to adopt new models easily without a prohibitively high cost of switching.

How can you choose specific cloud vendors while keeping an eye towards a more open, flexible future? One best practice is to pick vendors that also embrace the non-vendor-lock-in philosophy. Instead, vendors that don’t try to lock you into their proprietary technology will work to earn your business every day by maintaining good service and introducing new features.

OpenTelemetry made it possible for customers to improve their observability practice without being locked into a specific observability vendor. New Relic recognized that OpenTelemetry represents the future of observability, so we embraced it and took an OpenTelemetry-first approach by reimagining our tech stack with OpenTelemetry. For example, we adopted the OpenTelemetry-based OpAMP protocol for our fleet manager, and we’re also adopting OpenTelemetry-based telemetry pipelines so customers can optimize their telemetry data.

Each of these decisions demonstrates our commitment to an open future that doesn’t lock customers into our technology—you should look for that same level of commitment with each of your cloud providers.

How can you choose specific cloud vendors while keeping an eye towards a more open, flexible future? One best practice is to pick vendors that also embrace the non-vendor-lock-in philosophy. Instead, vendors that don’t try to lock you into their proprietary technology will work to earn your business every day by maintaining good service and introducing new features.

Preparing for challenges

Adopting an open-source, multi-cloud strategy is the right choice for the long term, but it comes with a set of challenges. Managing multiple cloud environments is complex, particularly when it comes to maintaining the systems and ensuring consistent security policies across platforms. Integrating open-source solutions into existing infrastructure also requires careful planning to avoid disruptions and ensure that the solutions meet the company’s needs.

To mitigate these challenges, companies should invest in training for their IT staff, adopt established best practices for multi-cloud management, and participate actively in the open-source community, which can provide valuable support and resources. By proactively addressing these challenges, companies can leverage the benefits of an open-source, multi-cloud strategy fully and turn potential obstacles into opportunities for innovation.

When industries transition from proprietary systems to open, standardized solutions, everyone benefits. These shifts lead to enhanced interoperability, giving users more freedom to choose between different products and services without being locked into a single ecosystem. The example of Apple’s transition from Lightning to USB-C enabled consumers to use a single type of cable for a wide range of devices, simplifying their lives and reducing electronic waste. The underlying goal of standardization—even something as complex as a multi-cloud strategy—is to enhance the user experience with more flexibility, better prices, and innovative services that meet their evolving needs.

For New Relic, our multi-cloud strategy gave us a competitive advantage and also positioned us to negotiate better terms with cloud providers. Our centralized full-stack observability platform aims to give our customers similar benefits, allowing them to manage their entire portfolio across multiple clouds. Most importantly, New Relic is the only observability vendor that can shift telemetry integration between clouds seamlessly, enabling customers to move between clouds easily without increasing their costs.

Open-source technologies offer a more user-friendly future with stronger product differentiation. We should do everything we can to develop and support those technologies.