4 Dominant Themes in the Discussion
| # | Theme | Key Take‑aways & Direct Quotes |
|---|---|---|
| 1 | Cheap, readily‑available server platforms | “It's an ancient old Dell PowerEdge R510… price of SSDs have skyrocketed so I'm not touching it.” – bombcar “HP ProLiant MicroServers (used). They support ECC ram, 4 caddies, one extra PCIe slot, and to my knowledge you're not CPU limited…” – burner420042 |
| 2 | ZFS‑specific failure handling | “zpool replace my_pool disk3 newdisk” – bobmcnamara “The guide isn’t complete without it – and the failure report should be active like a beep or flashing hardware light.” – beagle3 |
| 3 | Drive cost & sourcing trends | “I bought Qty 10 of HC520 (12TB) SAS drives for $1000… avg age is about 2.5 years, still well within its rated lifetime.” – neilpanchal “I ended up on shucking 4x the 14 TB WD Elements Desktop. They contain helium drives… about a third cheaper than 4x the 12 TB WD Red Plus drives.” – Confiks* |
| 4 | ECC‑RAM myth & ZFS stability across OSes | “ECC is better than non‑ECC, also for ZFS. The myth … is that ZFS requires ECC.” – magicalhippo (linking to a ZFS core‑developer rebuttal) “FreeBSD ZFS support has matured and is outstanding… the actual filesystem code is the same (and probably a bit more robust).” – ggm |
Summary
- Users gravitate toward inexpensive used enterprise hardware (Dell, HP, MicroServer) because of abundant drive bays and ECC support.
- ZFS users stress that simple zpool replace isn’t enough; proactive alerts (beep/LED or monitoring) are essential for reliable drive swaps.
- The market for high‑capacity drives is pricey; shucking and buying decommissioned SAS drives can be far cheaper than new consumer disks.
- The long‑standing claim that “ECC RAM is mandatory for ZFS” is largely a myth; ZFS works fine on non‑ECC RAM, though ECC is still preferable, and ZFS stability is comparable across Linux and BSD when using the same OpenZFS code base.