Deploying Databases on Kubernetes: Helm vs. Operators - Eric Schäffer and Manuel Swiercz

on YouTube ↗27m

" summary

Teams deploy Postgres and Elasticsearch daily using Helm charts and operators, and recommend Helm for packaging and composition while using operators for full lifecycle/autopilot where needed. Operators can implement capability levels up to 5 (autopilot, tuning, scaling); Helm covers levels 1–2 (install/upgrade). OperatorHub lists ~325 operators; Helm sees hundreds to thousands of chart releases monthly. Real deployments combine Helm to install operators plus cronjobs, secrets, metrics and backups; Elastic operator causes the most friction and production use is still being validated.

AI-written summary · about

" chapters (14)

1. 0:00Introduction and speakers
2. 0:44What is a Kubernetes operator
3. 2:27Components of an operator
4. 3:32Operator control loop explained
5. 4:59Operator advantages for databases
6. 6:46What is Helm
7. 7:51Helm advantages and use cases
8. 9:37Comparing Helm and operators
9. 11:27Capability levels and community support
10. 13:56Helm vs operator myths clarified
11. 15:02When to choose Helm or operator
12. 17:30Conclusion and real-world use case
13. 19:41Credits and thanks
14. 20:22Q&A session begins

" related talks

• 47mThe Automation Challenge: Kubernetes Operators vs Helm ChartsAna-Maria Mihalceanu · Jfokus 2022
• 32mThe Automation Challenge: Kubernetes Operators vs Helm ChartsAna-Maria Mihalceanu · Voxxed Days Ticino 2021
• 29mHelm Your Way with KubernetesAna-Maria Mihalceanu · GOTO Amsterdam 2022
• 30mThe Automation Challenge: Kubernetes Operators vs Helm ChartsAna-Maria Mihalceanu · GOTOpia February 2021
• 26mHelm your way with KubernetesAna-Maria Mihalceanu · Voxxed Days Microservices 2019
• 40mCustom Microservices Deployments Using Helm OperatorsDouglas Lopez · KCD Guatemala 2023
• 35mOptimizing Kubernetes Deployments with HelmErwin de Gier · GOTO Berlin 2018
• 41mHelm - the Better Way to Deploy on KubernetesReinhard Nägele · Codemotion Amsterdam 2018