Proxmox Advanced Training Feedback & Highlights:

Summary

https://youtube.com/playlist?list=PLZaEVINf2Bq_rQNlYDruK_vWrKU7N6Dn2

Notes

"People who don't take the training don't know what they don't know"

"If you're thinking about purchasing support, think about taking the training first - before you buy the hardware, if you can. It'll save you money, and make the time you spend on support much more valuable" (paraphrased)

"Install" vs "Advanced" is more "Virtualization" vs "Clustering" - it's all advanced
Weehooey's approach
- Decision-oriented, Experience-oriented
- labs didn't have a single hitch, everything was provisioned perfectly
Web UI + CLI + Important Locations, I.C.E.
Cluster install
- qdevice
- 2-4 nodes, more oversight
- 7+ nodes, easier
PBS
- block-based, live restore, metadata devs
- hourly backups, daily pruning
- needs special SSD dev if HDD
Corosync troubles
- 5-7ms latency
- dedicated / prioritity nic
ZFS vs Ceph
- ZFS > LVM + ext4
- speed vs parallelism
- high-endurance drives (high DWPD, even for OS disk)
- lower backfill warnings
- Proxmox vs DIY
- io threads, discard, qemu agent, fstrim
- bonded 10g, 25g, 100g
SDN
- self-documenting VLANing
- simple VxLAN EVPN BGP
HA 99.999% uptime (<4min down / year)
- shutdown policy = migrate
- qm create <vmid> to test strategies
- ceph: hyperconverged (4+ node)
- zfs: replication (2-3 node)
API
- permissions
- pvesh as an onramp from clickops to devops
PDM
- at-a-glance, live migrations across (CLI for now)

Resources

Proxmox Module Outlines https://www.proxmox.com/en/services/training-courses/training
Weehooey's Proxmox Training https://weehooey.com/pages/proxmox-ve-training
Root https://therootcompany.com

coolaj86/proxmox-cluster-course-feedback.md

Summary

Notes

Resources