Day 2 Operations Best Practices
summary
Yugabyte Platform is an enterprise control plane that provisions and manages clusters across clouds with API-driven automation. Demo kills one VM and the cluster auto-fails and rebalances tablet shards (default eight shards/node), restoring consistency after the shortened 60s failure timeout; live workload shows hundreds TPS on ~13 GB test data and recovery without app downtime. Customers run large clusters (examples: ~9 nodes across 3 regions, replication factor 3; one customer reports zero downtime for 300 days and another has ~2 TB per node). Backups go to NFS/S3, metrics feed Prometheus, alerts pair with PagerDuty, and TLS/KMS encryption is supported.
AI-written summary · about
chapters (26)
- 0:00Introduction
- 1:10DBRE Model
- 2:55Cloud Native Definition
- 4:14Yugabyte Platform
- 5:39SelfService
- 8:32Topologies
- 10:24Automation
- 11:00Data Recovery
- 13:10SelfHealing
- 13:33Configuration Management
- 14:12Monitoring
- 16:03Security
- 16:57Recap
- 17:14Live QA
- 17:50Cluster Design
- 18:21Platform Overview
- 19:09Platform Recap
- 20:05Universe Configuration
- 21:29Workload
- 22:08Nodes
- 22:57Leaders
- 23:26Delete Instance
- 24:40Check TServer
- 25:47Rebalancing
- 27:35Remove Node
- 29:25Create New VM
related talks
29mEnterprise Scale: How a Distributed Architecture Can Deliver Scale and Resilience
57mNo Downtime, No Limits
28mAutomation: Work Smarter, Not Harder. Using Valuable APIs, Capabilities & Ecosystem Tools
25mDiscover the Easiest Path to Distributed DBaaS: Yugabyte Cloud
34mWhat’s New: Discover the Latest Innovations and Updates from Yugabyte
32mZero-Touch Fault-Tolerance for Cloud-Native Geo-Distributed Databa...
01h 29mYugabyteDB DBA Fundamentals
30mBuilding Cloud Native Applications on Distributed SQL Platform