Database sharding at the load balancer level

on YouTube ↗52m

" summary

Describes how T-Bank shard a high-load customer-data service with tables of billions of rows by implementing sharding at the load-balancer layer in Kubernetes. Covers the technical challenges encountered during that rollout and the solutions used to deploy the change with minimal user-visible impact.

AI-written summary · about

" chapters (8)

1. 0:00Introduction
2. 4:43Vertical scaling problems
3. 7:07What is sharding
4. 11:19Sharding key choice
5. 16:56App problems with sharding
6. 21:06PostgreSQL replication
7. 27:09Shard launch and migration
8. 31:34Q&A

" related talks

• 22mScaling a Financial Institution with KubernetesYago Nobre · DevOpsDays Vitória 2020
• 32mSharding for mission-critical, internet-scale appsShailesh Dwivedi · Oracle DatabaseWorld 2021
• 46mDatabase Sharding the Right Way Easy, Reliable, and Open sourceEsen Sagynov, Lee Seongjun · HighLoad++ 2012
• 23mSharding for hyperscale and geographically distributed relational databasesPankaj Chandiramani, Shailesh Dwivedi · Oracle CloudWorld 2022
• 27mAutomatically shard & scale-out traditional databases on K8sTrista Pan · Conf42 Cloud Native 2023
• 35mIntelligent Load Balancing in KubernetesGaurav Nanda, Vincent Cheng · SREcon26 Americas
• 29mCloud Native Patterns – Shifting ResponsibilitiesCornelia Davis · Distributed SQL Summit Asia 2021
• 31mBeyond the Limits: Scaling Kubernetes Controllers HorizontallyTim Ebert · KubeCon + CloudNativeCon Europe 2025