Azure & DevOps Podcast

Martin Thwaites: Building Software for Production - Episode 271

Nov 13, 2023

Martin Thwaites, Developer Advocate at Honeycomb, discusses observability in software production, tracing unknown unknowns, open telemetry, and the mechanics of committing to the main branch in software development.

42:47

Creator website

Episode guests

Martin Thwaites

AI Summary

AI Chapters

Episode notes

Podcast summary created with Snipd AI

Quick takeaways

Observability focuses on understanding the customer experience rather than technical metrics in order to build reliable software for production.

OpenTelemetry provides a standardized and vendor-neutral way of capturing telemetry data, making it easier to integrate observability into applications.

Deep dives

Observability and Building Software

Observability is crucial for building software that performs well in production. It involves understanding how our system performs and how it affects users. This goes beyond monitoring metrics like CPU usage and focuses on the customer experience. Reliable software should prioritize the impact on customers rather than technical metrics. To achieve observability, we need to ask questions that we didn't know we needed to ask. This requires adding contextual data to our production systems, such as user location, feature flags, and specific user actions. By adding context and asking the right questions, we can gain insights into production issues and address them effectively.

Introduction

2min

Understanding Observability and APM Monitoring Metrics

15min

Getting Started with Observability and Open Telemetry

4min

Adding Context at Application Edges for Enhanced Observability

5min

Believing and Empathizing with Users and Deploying Software to Production

2min

Committing to the Main Branch and Continuous Integration

15min

Martin Thwaites is a Developer Advocate at Honeycomb, an o11y enthusiast, and a delivery-focused Developer from the UK. With over 20 years of experience in development in the .NET ecosystem, he’s worked with many companies on scaling up engineering teams and products. The past few years have been spent working on solving complex problems with some of the UK’s big names, including e-commerce retailers and credit lenders.

Topics of Discussion:

[2:42] Martin getting his start in testing.

[5:55] What other products is Honeycomb similar to?

[5:57] APM monitoring metrics.

[9:05] O11y

[9:30] The foundational elements of Honeycomb.

[13:36] For smartphone applications, desktop, or mobile, what of these concepts are the same, and what’s different?

[15:49] Tracing the unknown unknowns.

[17:43] Where open telemetry comes in and shines.

[28:04] Do you commit locally, group them up together and execute a push?

[33:24] Moving TFS Code Base onto Git.

[34:40] What TFS did right.