DevOps Tools Evolution

This deck was presented to the DevOps Meetup on July 20, 2015.

Transcript

1. DevOps Tools Evolution
2. Today's Menu • AppsFlyer in numbers • Rnd structure in AppsFlyer • Our DevOps philosophy • Motivation for creating DevOps tools • Setting requirements and project assignment • Example for tools we built internally • Live demo and the developer angle
4. AppsFlyer in numbers • ~ 4.9 billions events per day • ~ 120 services • Elastic hundreds to 1K instances • Dozens of deployments per day • 25 people in development • 1 team
5. One team philosophy
6. That's me
9. Motivations • Actual operational needs • Extend existing tools • Learn new technologies • Familiarize the production environment for new Rnd members • Blur the line between Dev and Ops • Improve the bonding inside the team
10. Considerations • Needs, why do we need this project? • Usability – How do I see myself using the project? • The target audience. • The developer's background (Front, Mobile, etc) • Will the developer be better familiar with production environment? • How are we going to maintain the project? • What are the time constraints? • Can we roll it out in stages? • Can we release it outside?
11. Today's Examples • Sniffer: Mapping tool (Artyom Shein) • Sensu Reports: Aggregated reports of Sensu (Avner Schwartz) • Bouncer: Authentication and Authorization (Reshef Mann)
12. Live Demo
13. And... Rock n Roll
14. Questions?