ZFS on Proxmox: Tips for Performance & Efficiency 🚀

Whether you're hobbyist or managing a production cluster, ZFS gives Proxmox administrators enterprise-grade storage capabilities: snapshots, compression, replication, and rock-solid data protection.

In this session, Allan Jude and JT Pennington will walk through practical ways to get the most out of ZFS in Proxmox. From tuning for performance to designing efficient replication, you'll come away with tested practices you can put to use right away.

What You’ll Learn:
- How to tune ZFS for performance in Proxmox environments
- Compression and caching strategies for different workloads
- Replication workflows that balance efficiency with resilience
- Common pitfalls and how to avoid them

Learn more: https://klarasystems.com

Timestamps
0:00 – Intro
1:07 – Overview
3:12 – ZFS vs Ceph: Design, Complexity & Scaling Trade-offs
10:14 – Unclustered Pair vs Centralized Storage Architectures
15:00 – Performance Tuning: Record Sizes, Amplification & Compression
18:27 – CPU Pinning, NVMe Queues & Metadata Caching Explained
23:53 – ARC, L2ARC & Prefetch Behavior in ZFS
28:42 – Replication, High Availability & Disaster Recovery Scenarios
34:10 – Bookmarks, Snapshots & Incremental Backup Strategies
40:22 – ZFS Block Size Changes (8K → 16K) & RAIDZ Efficiency
47:09 – Hosting VMs on HDDs: Metadata VDEVs & SLOG Performance
49:08 – IOPS Bottlenecks & NVMe-over-Fabric Design Guidance
50:36 – Storage Hardware Tips: Drive Choices, Pool Space & SLOG Mirrors
55:22 – Closing Remarks, Poll & Upcoming Webinars