Sometimes I feel like I’m in my first Open Source Anonymous meeting. Hi, my name is Chris Lentricchia, and I haven’t used open source software for six months.
I’m 27, and Instana is my first proprietary software company. After working my entire career in open source, it’s pretty surreal.
I started at Red Hat, which was also “born” in 1993, where I oversaw an array of technologies, from Linux to management software and middleware. At SUSE, I oversaw product marketing for their Kubernetes platform. So I knew very little of proprietary software. In fact, if you had asked me six months ago if I wanted to work in closed-source software, my answer would have been a resounding “Hell no!”
But here I am.
Thanks for sharing, Chris.
I’d like to talk about the differences between Instana’s proprietary approach to observability and open-source solutions. I’m not trying to convince anyone that open source isn’t valuable or that one way is better than the other. They’re just different. So here are my observations as a new employee, with an open source background, learning a new type of technology.
Prometheus and Grafana: the Gold Standard for Open-Source Monitoring and Observability
There are a lot of choices out there for monitoring and observability. If you’re in the cloud-native ecosystem, you know the “Cloud Native Hellscape” I’m talking about. But when we’re talking monitoring and observability, Prometheus and Grafana are really the gold standard. Many Kubernetes distributions, like the one I worked on at SUSE, ship with Prometheus and Grafana for observability natively. So today, I’m using Prometheus and Grafana as my open-source monitoring and observability examples.
Prometheus and Grafana are two great pieces of software that are commonly paired together to create a monitoring solution. Depending on your familiarity with cloud-native technologies in general, they can be a convenient tool for learning about the APM space, as they can be easy to get up and running. Of course, nothing in software is as easy as marketing would have you believe.
Prometheus and Grafana are pretty widely used across a variety of use cases and ecosystems, from a single person monitoring her personal home power consumption to large telecommunications companies. They do require some configuration and often require other pieces to complete the observability puzzle – sometimes with very nice results! Many open-source companies ship Prometheus and Grafana with some pre-configured dashboards and setups to get users up and running. These dashboards may not cover your specific use case, but you can modify them to fit.
Tell us more Chris.
Grafana for open-source visualization
Grafana is an open-source solution used for metrics, visualization, monitoring, and analysis. It is developed for broad compatibility and boasts an active developer community for both support and rapid innovation. Grafana offers an amazing selection of features and setups. However, those features often require heavy customization to get up and running. Grafana setup may require the installation of plugins, which can also be a bit finicky.
As my introduction to the world of monitoring, Grafana proved to be highly valuable. It’s free, there is little barrier to entry, and it’s highly customizable. It allowed me to tinker around with a myriad settings and dashboards – basically until I was bored, frustrated, or both. Finally, Grafana includes support for Prometheus, it’s commonly paired twin.
Prometheus for time-series data
Prometheus is an open-source systems monitoring and alerting kit that displays time-series data. Per the Prometheus website, Prometheus is designed for reliability, meaning it will be around when your other systems go out. You don’t have to set up a massive amount of infrastructure to use it. Prometheus supports traditional and highly dynamic environments.
Prometheus also boasts a weighty developer community for both innovation and support. Again, being open source, Prometheus provided me with an excellent playground where I could become dangerous in the monitoring and observability game. That said, nothing comes without drawbacks.
Prometheus doesn’t provide a dashboard (that’s what Grafana is for). Most important for this discussion, Prometheus requires a healthy background in IT. It has a default retention time of 31 days and the ability to scale only vertically. There are other projects to address those shortcomings, like Thanos for instance, but they are out of scope here.
Would you like to talk about proprietary solutions?
Open source comes from a mindset of freedom and customization. Proprietary solutions may not be as customizable as some open-source solutions, but they have their own distinct advantages. Some focus more on being “plug-and-play,” while others build on reliability.
Proprietary solutions are also on an open source sliding scale; some use elements of open source, and some don’t use any at all. Just like open-source solutions, proprietary software offers a variety of solutions with distinct advantages and disadvantages. I’m using Instana as my example because, as discussed, it’s the only closed-source tool I’ve worked with.
Instana for automated observability
Instana offers a take on observability that really intrigues me as someone without a background in IT. Among other benefits, Instana offers automation, and its platform requires little to no prior knowledge or effort to set up and get running. This approach interests me because of my background, and because it brings efficiency to the observability game.
Upon installation, Instana immediately and automatically connects to all the end points in a given ecosystem. And that doesn’t just equate to an easier learning curve, either. It also means that developers don’t need to worry about writing observability into their applications. Instana offers a simple “one stop shop” for observability. While still being supported, Instana offers a complete package for the observability space that doesn’t require much button pushing action – which is right up my alley. In my view, and in a purely theoretical sense, Instana takes what I already knew to be great with open source monitoring platforms and builds upon it.
Instana is also open-source friendly, so it’s not a complete departure from my wheelhouse of open source. Many customers pair Instana with OpenShift and VMware Tanzu. Instana also utilizes some open-source components in the creation of its platform. It’s a tool that I would use if I were going to be using an observability platform on a daily basis. That might seem simplistic; but let’s face it, we don’t all endorse the products we work on.
Turns out there are actually a lot of ways to eat the proverbial elephant. Instana tackles observability from a slightly different angle than open-source tools like Prometheus and Grafana. While open source usually comes from an angle of customization, Instana “just works right out of the box.”
Neither approach is necessarily better; but after years working with platforms built for customization, I was ready to work on one that just worked. As noted, Instana does work in close proximity to popular open-source tools and even has some open source bits baked into it. Altogether, it isn’t so much of a departure that it offended my open source sensibilities, but it does offer some interesting advantages that I was ready to sink my teeth into.
All in all, Instana offers a pretty cool tool. If you’re still reading and you haven’t fallen asleep, feel free to check out Instana’s online sandbox, the Play With App, or use it free for 14 days, to try it yourself!
Oh, and coffee and doughnuts are in the back of the room.