Guest: Asaf Yigal (LinkedIn)
Company: Logz.io (Twitter)
Show: TFiR: T3M

Logz.io is a visionary when it comes to application performance monitoring and observability. In this episode of TFiR: T3M, Asaf Yigal, Co-Founder and CTO of Logz.io talks about the current trends and challenges in the observability space.

Highlights of this video interview:

• Observability is the ability to measure how a company’s technology is impacting the business, to measure service level objectives (SLOs) and be alerted when service is not within the objectives that have been set.
• Monitoring is keeping track of objects such as a server, Kubernetes cluster, code, etc. and making sure they’re running and they’re not reaching the limits. Observability is tying all the pieces together. If you have an objective, e.g., you want your purchasing process to be less than a second or you want your purchasing process to have an error rate of less than 0.01%, you can set up observIQ, set up goals and objectives, and you can basically create observability. The underlying components would be logs, metrics, and trace information.
• Kubernetes is an amazing open-source service that everybody’s using, but it creates challenges when you want to create observability.
• The more flexibility you have in your environment, the more layers of abstraction, the more you need observability.
• With so much data to sift through, organizations are now more concerned about the cost of the cloud as well as the cost of observability.
• A major challenge for all the observability vendors is how to collect the information simply and quickly. With Kubernetes, there are many layers of abstraction. There is a need to collect the logs of the cluster, the logs from the actual nodes that are running, etc. The entire industry joined forces to solve that problem with OpenTelemetry. Per Yigal, it is definitely making strides in the right direction and hopefully going to mature over the next two to three years.
• Companies should care about observability because there is no other way to achieve business agility.
• Most of the observability, especially around the open source, requires knowing a specific language like PromQL to query data and knowing what to commit. Generative AI may be able to construct the right query and list all the problems or highlights of the problem in my environment. For example, “Give me all the servers that crashed last night.” No need to know how to construct that query. Generative AI is an opportunity, rather than a threat in the observability space.

Logz.io aims to make life simpler for engineers, DevOps, and executives in companies by giving them:

• The ability to shift from monitoring to observability
• The ability to define service level objectives (SLOs)
• The ability to view their process end to end, with all its underlying microservices.

Advice for companies looking to start their observability journey:

• Set objectives: what do you want the observability to report to you?
• Understand the business impact.
• Invest in data collection and data hygiene. Many organizations just collect everything they can get their hands on. It costs a lot of money, creates a lot of clutter in the environment, and hampers the ability to search and gain insights. Be very detailed about the data that you collect, where you’re going to collect it, and how it is going to tie to the rest of the data you already have.

This summary was written by Camille Gregory.