How to Run Your Own Gitlab Runners as Your Company Grows

In this session, Casey Zednick, Senior Software Engineer for NGINX, tells you everything you need to know about GitLab Runner, an app for running jobs in your CI/CD pipeline. Using your own runners gives you full control to meet your service level objectives (SLOs), increase security, and save money at scale. It's notoriously difficult to forecast costs for dynamic cloud services like an autoscaling fleet of runners, but Casey demos a tool from the NGINX tools team that calculates your costs based on a handful of numbers. He also shows how to clone our GitLab repos to set up a production-grade fleet in Azure and monitor its performance.

Speakers:
Casey Zednick

How to get started with NGINX

The Complete NGINX Cookbook

Chapters
00:00 Introduction
00:32 What is Gitlab?
00:55 What are Gitlab Runners?
01:33 Why run your own runners?
03:40 Understanding costs of running a fleet
04:22 Forecasting Fleet Costs
06:26 Optimizing Fleep Compute Costs
07:57 Understanding the Key Parts of Fleet Design
08:25 Hosting the Runner (Control)
09:18 Understanding Autoscaling Executors (Compute)
12:15 Establishing Service Level Objectives (SLOs)
15:28 Gathering Service Level Indicators (SLIs)
16:56 Visualize Service Level Indicators (SLIs)
18:55 Resources