Shelby Spees

Developer Advocate at Honeycomb

Los Angeles, California, United States

Shelby Spees is a Developer Advocate at Honeycomb, where she helps developers better understand their services in production in order to deliver more business value. Before joining Honeycomb, Shelby worked on applications, build pipelines, infrastructure, and other reliability efforts. She’s dedicated to improving access and equity in tech as well as to helping people feel good about their work.

Shelby lives in Los Angeles, CA, where she enjoys doing karaoke with her pitbull, Nova.

Area of Expertise

Information & Communications Technology

Topics

Observability
devops
Distributed Software Systems
Site Reliability Engineering
Service Level Objectives
DevOps & Automation

Observability for Software Teams

Observability, instrumentation, telemetry--what does it all mean? This introduction to observability is for software practitioners who want to better understand the health of their production systems. Learn how to generate better data and gain new insights. You'll walk away ready to use observability to level up everyone on your team!

Think Prod Up Front: Shifting Left on Production Excellence

We all want to be careful, but not everyone knows how apply “be careful” effectively. There’s good news, though: we have the technology! Learn how application and platform teams can leverage modern software delivery tools to shift left on service reliability while reducing stress and delivering better business value.

Getting Started with Observability for Chaos Engineering

How much are you learning from your chaos experiments? How do you know it’s safe to perform them? Learn how to instrument your systems in order to gain confidence and get the most out of chaos, whether you’re injecting it on purpose or it’s already there in your system. This talk explores how Honeycomb’s engineering team leverages high-context telemetry data to measure service heath and risk in order to determine when it’s safe to perform experiments. Attendees will come away with next steps for instrumenting for observability in order to get the most out of their own chaos experiments.

Case Study: Experiments on Honeycomb’s Stateful Ingest Service

This talk is a deep-dive discussion of the experiments Honeycomb’s platform engineering team performed on the stateful parts of Honeycomb’s backend data ingest service, which uses Kafka and Zookeeper. Hear about lessons learned and how the team gained the confidence to make an architectural change that resulted in a 40% reduction in AWS spend for that fleet.

Graviton2: How Honeycomb Reduced Infra Spend by 40% on Its Highest-Volume Service

How did Honeycomb have the confidence to migrate to a more efficient processor architecture, resulting in a 40% reduction in AWS spend on their backend ingest service? In this talk, Developer Advocate Shelby Spees explores how Honeycomb’s platform engineering measured the risk and ROI of moving their high-traffic ingest workloads to the new ARM-based Graviton2 processor architecture. Learn how high-cardinality observability data supported performing side-by-side comparisons of the same workloads on different processor architectures, and how to gain similar insights into your own infrastructure.