Lessons scaling operations to everyone @indix

on YouTube ↗17m

" summary

Indix moves from one ops person and a centralised chef server to teams owning infrastructure with Ansible, Marathon/Mesos and GoCD; engineering headcount grows from ~5–15 to 30–50 while ops stays ~1–3. Claims: decentralisation reduces midnight pager helplessness but drives cost spikes ($150k→$220k/month) and fragmentation. Concrete fixes: per-team playbooks, operability review checklist before production, DNS backups every 30 minutes after a production overwrite, and a tech radar to standardise tooling.

AI-written summary · about

" chapters (11)

1. 0:00Intro
2. 1:05Three stages
3. 1:19Early stage
4. 4:56Devs
5. 6:09Growth
6. 8:25Cost
7. 10:58Later stages
8. 12:39Microsoft schedulers
9. 13:54Tech Radar
10. 15:06Operability Review
11. 16:10Recap

" related talks

• 52mDevops @ScaleBaruch Sadogursky · Devoxx Belgium 2017
• 48mDevops@ScaleLeonid Igolnik · GeeCON 2019
• 55mDevOps @Scale (Greek Tragedy in 3 Acts)Baruch Sadogursky, Leonid Igolnik · JavaDay Ukraine 2017
• 58mDevOps @ Scale — Greek Tragedy in Three ActsBaruch Sadogursky, Leonid Igolnik · DelEx Conf 2018
• 36mDevOpsSup - Extended DevOps approachGrzegorz Sztandera · Conf42 DevSecOps 2023
• 18mScaling DevOps in Complex Environments: Challenges, Failures, and SuccessesChaitanya Rahalkar · DevOpsDays Austin 2025
• 57mGo Away Or I Will Replace You With A Very Little Shell ScriptKristian Köhntopp · FrOSCon 2015
• 22mRaise a Production System:Making on-call human-scale w/ Observability & TrustChris Sun · ScaleConf Colombia 2019