Virtualization Technology News and Information
Rookout 2022 Predictions: Observability evolves from static to dynamic

vmblog predictions 2022 

Industry executives and experts share their predictions for 2022.  Read them in this 14th annual series exclusive.

Observability evolves from static to dynamic

By Liran Haimovitch, CTO and Co-Founder of Rookout

Observability is one of the hottest topics in software development -- and for good reason. As digital transformation continues to accelerate for organizations around the world, which by extension means rapid adoption of modern cloud architectures that are agile and scalable but also highly dynamic and complex, the tooling required to derive insights and maintain the health and stability of these applications is evolving quickly.

It wasn't long ago when the majority of organizations operated their own infrastructure. The idea of offloading computing and storage to a 3rd party like AWS was so revolutionary that it can't be overstated. While traditional monitoring and APM tools from the big four management vendors BMC, CA, HP, and IBM attempted to build cloud-based solutions for better visibility, the fact of the matter is cloud-native applications required a new approach.

As of writing, Datadog is a $50Billion dollar company. It's core product dating back to the launch of the company in 2010 was a hyperfocus on cloud infrastructure monitoring, providing dashboards and alerts when certain thresholds were crossed that required attention. Many other traditional cloud monitoring solutions such as New Relic and Dynatrace have seen a great deal of success as well, and the fact of the matter is traditional monitoring isn't going anywhere, anytime soon.

But in the world of microservices, containers, and highly ephemeral applications, it's become impossible to fully understand the complexity of a modern application well enough to predict in advance everything that needs to be alerted on. And with the rise of automation, change intelligence, and AIOps, it's becoming more and more true that if a problem can be predicted in advance, it should also be automatically resolved. Humans should only be paged and woken up in the middle of the night for truly novel situations.

This is where the world of Observability comes into play.


Thanks to companies like Honeycomb and Lightstep, we're seeing the benefits of being able to ask real-time questions of our systems, without having predicted in advance what questions were worth asking. Engineers can now easily slice, dice, and query data with high amounts of telemetry to quickly observe the current status and health of their applications. To many of us, this novel approach still feels like a superpower.

But where can we go from here? As I see it, the limitation of Observability is that while you can ask real-time questions, the answers you get are still static. With traditional observability tools, when you want to have access to a specific metric, trace, or other piece of information about your application, it often requires you to make a new configuration change, code change, or even a redeployment of your application. If you couple this requirement with enterprise release management practices, where environments are more heavily regulated and often require one or more approval gates, this can dramatically slow down the troubleshooting process.

The future is Dynamic Observability, which in short means that teams should have access to gather any data point within their application in real-time AND on demand. To accomplish this at Rookout, we are doing bytecode manipulation that enables developers to extract data from production systems in real time. If a developer wants to know the value of a variable in their code, for example, and they don't yet have a log line or a metric that gives them insight into that data, with Rookout they are able to collect that information as it's needed without having to make a change to their code and without having to redeploy their application or service.

This is the future: click a button, get the data.



Liran Haimovitch 

Liran is the Co-Founder and CTO of Rookout. He's an advocate of modern software methodologies like agile, lean, and DevOps. Liran's passion is to understand how software actually works.

Published Wednesday, November 17, 2021 7:37 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<November 2021>