Managing Data at Scale: Best Practices and Evolution of SIG-Apps
Talkat KubeCon + CloudNativeCon Europe 2025 · on Friday, 2025-04-04· at 13:45· Level 3 | ICC Capital Suite 10-12·
summary
SIG Apps reports concrete changes: pod disruption budgets gain a field to count ready-only PDB counting so nonready pods can be evicted; ReplicaSet downscales now pick pods more randomly; StatefulSet adds startOrdinal for StatefulSet and explicit PVC policies for scale-down vs delete; Jobs add elastic indexed jobs with stable indices, annotations for cronjob creation time and pod index, a per-index backoffLimit (default global backoff is 6), and a job success/completion policy. Node maintenance/evacuation and job pod failure policy proposals remain in active design. Meetings every other Monday on Slack.
AI-written summary · about
chapters (22)
- 0:00SIG-Apps introduction
- 1:04Engaging with SIG-Apps
- 2:06SIG-Apps core controllers
- 3:13Annual reports and historical evolution
- 4:20Stable features past year
- 5:25Pod disruption budgets
- 7:13ReplicaSet downscaling
- 8:18StatefulSet migrations
- 9:21StatefulSet PVC policies
- 11:10Batch Working Group
- 11:53Elastic Index Jobs
- 14:02CronJob delays
- 15:06StatefulSet annotations
- 16:13Kubernetes GA features
- 16:33Job backoff limits
- 17:37Job completion policies
- 18:20External job controllers
- 21:35Job failure policies
- 22:40Deployment rollout pacing
- 24:08Node draining
- 25:57StatefulSet challenges
- 29:08Q&A
