Skip to Content

Observability for Sustainability

Mark van der Walle
March 08, 2023

The battle against climate change is one that we cannot afford to lose. It is a battle in which all of us are enlisted to make a difference.

The practice of doing IT comes with an unfortunate side effect. The footprint of digitization is enormous. Whether this is about on premise data centers, cloud or workplaces in the office and even at home. This brings a large responsibility of tackling climate change in this area to those who make IT their profession.

We as IT-professionals are uniquely positioned to make a huge impact on society through the application of green IT.

A cloud for everyone

It would be obvious to just say: “Go to the cloud!”or “Use serverless!”. And those are indeed routes to explore. Public clouds are far better at consolidating resources and using serverless allows organizations to use the cloud truly on demand. However those routes are not always applicable or feasible for many differing reasons.

So what more can we do? If we look at tackling climate change there are a few approaches. One of those is to compensate for our carbon footprint by planting trees and other such compensating approaches. Another one is to use sustainable energy sources. And yet another is to reduce the amount of resources we use. This last one is where going to the cloud and using serverless are oriented.

Beyond the obvious

As mentioned earlier, going to the cloud is not always feasible and beyond that we are still dealing with infrastructure inside offices or at home. So what more can we do? What if we could make our applications more efficient so they use less resources? This could lead to scaling down our infrastructure. How would we do that?

Enter observability

To know where to start requires knowing where the most impact can be made. We can do this by observing how our applications behave in real world situations. This leads to the ability of using a data-driven approach towards making a significant impact on the efficiency of our landscape.

Let us do a recap of what observability means within an IT landscape. Observability is the ability of perceiving state and behavior of IT-systems at scale. It allows you to answer complex questions about your applications, infrastructure and the interactions between them. Observability is based on the three pillars of logs, metrics and traces. There could be some variations, but in general everything in the observability landscape can be placed into those three categories.

How does this help?

So how does observability help? Observability does not only show you that CPU usage is high, but you can also get answers as to why that is happening. And if you know why you can take action. This can be either to solve a problem, but also to improve efficiency.

Let’s make this more clear with an example. Imagine we have an application where a user can request specific information. Now imagine also that this information requires intensive computation to get to a result and also requires the application to do numerous requests to a database. Now imagine that the application is running fine so there is not much reason to look at it.

Except that this application is probably using lots of resources to get this done. Now by clearly examining the data of when a user requests information and the resulting amount of effort by the application, database and network infrastructure we hopefully come to the conclusion that we can do better.

We could for example cache parts of the database accesses or parts of the computation or perhaps even the entire result depending on the scenario. This will not only lead to reduced resource usage, but also increase user experience and reduce operational costs. And perhaps we can scale down our infrastructure.

There is a cost, but we should pay it

Observability is not a silver bullet and there is a cost to applying it. We need to store more data and perform more computations to get proper insight. Also there is always a tradeoff between putting data in memory, on disk or to compute it each time.

However the benefits that observability brings to the table for solving problems, making better decisions and especially tackling climate change far outweigh those costs. Observability is here to stay and it is not for nothing that Gartner has named Applied Observability as one of the 2023 Technology Trends.

Observability for Sustainability, our responsibility should not be anything less.

About the author

Lead Software Architect | Netherlands
Mark is an experienced software architect with more than 14 years professional experience in software development and operations. In his years of experience Mark has always had a drive of building and designing reliable and simple solutions to complex problems. To realize this Mark has a strong focus on quality backed by solid engineering, CI/CD pipelines, DevOps principles and craftmanship.

    Comments

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    Slide to submit