AWS Open Source Blog

How AWS and Grafana Labs are collaborating to improve Grafana for all

Torkel Ödegaard was fed up. As a developer and architect, he says he struggled to get people on his team to instrument their applications and services and build dashboards to make analyzing and understanding the company’s application data over time easier. In our recent interview, Torkel says his former employer had a forward-looking microservices architecture, but that required better observability of the system than they had. So he started to tinker on a hobby project on the side, and that tinkering resulted in Grafana. After a few months working early mornings, late nights, and weekends to turn his dashboarding dream into a reality, Torkel quit his job so he could work on Grafana full-time and see where it would take him. It was a leap of faith, with no clear path to turn growing industry interest into income.

That was 2014. Fast forward to 2020 and Grafana has become the industry standard for visualizing observability data, with users as varied as JP Morgan Chase, Bloomberg, PayPal, and Verizon. In fact, hundreds of thousands of companies run Grafana today, with installations roughly doubling every year for the past several years. Torkel says this success is a bit mind boggling, but also gratifying. Still, Torkel and the company he co-founded, Grafana Labs, are constantly iterating on ways to fund growth for the Grafana project. This is one big reason Torkel was optimistic when AWS reached out in 2019 to him and his co-founder and CEO, Raj Dutt, to discuss a managed Grafana service.

Although the companies had collaborated for some time on Grafana data source plug-ins for a variety of AWS services, how the two companies would work together was not immediately clear, especially given Grafana Labs’ own plans to run a managed service. AWS support for Grafana would help to further validate it as an industry standard, Torkel suggests, but was there room for two managed Grafana services? For Raj and Torkel, any concern was tempered by two things. First, Raj explains, they decided “it would be better to be in the boat with AWS,” building a service together. Second, there was optimism that this could be win-win.

This is the story of how AWS and Grafana Labs are working together to ensure the long-term success of Grafana.

Grafana’s big tent philosophy

In late 2014, Torkel connected with Raj and Anthony Woods, Grafana Labs co-founder and CTO. The three had been wanting to build something in the monitoring space using open source, and in early 2015 they founded Grafana Labs to provide products and services around Grafana. What those products and services should be, however, remained a question. For the first three years, the company focused exclusively on the open source project before they introduced any enterprise differentiation as a way to drive subscription revenue.

Over the years, the company has tinkered with different business models. The original business model called for a SaaS monitoring platform. Later they realized there was big demand for an on-premises product, as well, which became Grafana Enterprise. “It took a long time,” Torkel remembers, but that enterprise differentiation eventually crystallized into a model by which they sold access to enterprise plug-ins for Splunk, Oracle, and other commercial data sources.

In this, Grafana Labs took a very different approach from many other projects and companies.

One thing that sets Grafana apart is its “big tent” philosophy, Torkel says. That philosophy shows clearly in the data sources that Grafana Labs supports. “We have data sources for many SaaS vendors that we consider competitors,” he says.

Although this may seem counterintuitive, Torkel explains, “We have a ‘big tent’ philosophy where we see the pie is big enough for everyone, and we want to integrate and make those users part of the Grafana family. That’s one of the key things in Grafana: a multi-data source architecture that gives people the ability to bring in data from different vendors and different open source projects.” He says that Grafana’s focus on the customer means that it supports data sources that might be in contention or in competition with Grafana in some ways.

Importantly, this approach enables Grafana to bring together the enterprise hodgepodge of different data sources, Raj says. He says that the value for customers is that if they buy Grafana, “We’ll help you make your very heterogeneous world make sense.” It’s also in this plug-in approach that Grafana Labs and AWS saw an opportunity to collaborate.

AWS and Grafana Labs: Better together

Developers and operations teams use multiple tools to collect and store different types of operational data for their applications. The variety of both AWS and third-party data sources makes getting a consolidated view of application health and performance difficult. Many enterprises depend on Grafana to query and visualize data across multiple data sources for metrics, logs, and traces. However, customers must manually provision, manage, maintain, and scale Grafana themselves, which adds operational complexity. Customers want the flexibility that Grafana provides for querying and visualizing all of their operational data, without the complexity of managing the setup, operations, and security integrations.

AWS wanted to offer customers a fully managed Grafana service, and felt the best way to do this was through a commercial partnership with Grafana Labs.

In 2019 AWS started working through ways to support customers with a Grafana service while also helping to sustain the success of Grafana Labs. At Amazon, customer obsession is the primary, guiding principle. That same customer obsession applies to partners like Grafana Labs. It’s part of the AWS mission to be the best place for customers to build and run open source software in the cloud, leading AWS to work closely with open source partners like Grafana Labs on go-to-market activities and, depending on the project, contributing through code, credits, or whatever the partner and its community may need.

As mentioned, AWS and Grafana engineers have closely collaborated for years on Grafana data source plug-ins for Amazon CloudWatch, AWS X-Ray, and Amazon Timestream—each open source and used daily by thousands of customers. But this went beyond engineering collaboration. This would be a business partnership.

As AWS and Grafana Labs worked through partnership options together, it became clear that this could be a “win-win” for both companies and our customers if we structured the partnership appropriately.

On the customer front, as fantastic as open source Grafana is, Grafana Labs has augmented Grafana with additional functionality that customers might like. AWS, too, has added to Grafana in the Amazon Managed Service for Grafana, deeply integrating Grafana with native AWS capabilities such as AWS CloudTrail; AWS Single Sign-on; multi-account, multi-region access; and auto-scaling. In these ways AWS and Grafana Labs are combining the best of open source with the best of cloud, serving our respective customers through our different managed services.

On the partnership side, one of the biggest requests from customers is enterprise plugins for proprietary data sources, such as Splunk, DataDog, AppDynamics, etc. To get easy integration with such data sources, Grafana Labs offers Grafana Enterprise, which AWS now offers as an in-place upgrade directly within the Amazon Managed Service for Grafana service console. For Raj, it’s this joint go-to-market that offers the most promise to make this partnership successful, and thereby help to fuel even more Grafana innovation.

Conclusion

AWS, working with Grafana Labs, will be contributing licensing revenue and code to help make Grafana even better, not just for the AWS service, but also for open source users and Grafana Cloud customers. Additionally, AWS and Grafana Labs are partnering to provide open source Grafana as a managed service to AWS customers. Finally, AWS will promote and sell Grafana Enterprise for customers who want access to proprietary data sources, thereby helping to further fund the innovative work Grafana Labs does for Grafana.

For Torkel, this collaboration is important. He says that only will this collaboration between AWS and Grafana Labs raise interest in Grafana, but the commercial arrangement should mean that both parties are going to benefit both commercially and in terms of contributing to the project. In these ways, we hope that our customers, the Grafana community, and the two companies will benefit, resulting in a better, more financially secure Grafana project.

Interested in helping out with Grafana? Check it out on GitHub and let’s work together to make it even better.

Matt Asay

Matt Asay

Matt Asay (pronounced "Ay-see") has been involved in open source and all that it enables (cloud, machine learning, data infrastructure, mobile, etc.) for nearly two decades, working for a variety of open source companies and writing regularly for InfoWorld and TechRepublic. You can follow him on Twitter (@mjasay).