4 ms·

Every time I make this argument (whether regular RAID or ZFS), everyone pulls out their pitchforks and tells me about how they have run RAID 5 or 6 for years an

Every time I make this argument (whether regular RAID or ZFS), everyone pulls out their pitchforks and tells me about how they have run RAID 5 or 6 for years and never had a problem, plus "I can have two drives fail! You would lose your array if the wrong two drives fail!" But drive failure is much more likely during an operation like a resilver/resync, and I wonder if the majority of people espousing more risky RAID a setups (especially single parity) are those who just want hundreds of TB for a growing media collection. I know for my critical data, I don't trust parity. Plus I can't afford the days or weeks long resilver operation, I need a working storage array that doesn't suffer drastically in performance if one drive goes bad.

Cyph0n 6y ago
Critical data for the average user typically does not require much space, so the cost of storage overhead in a mirrored ZFS pool is not too bad. RAIDZ is much more cost efficient when you need to reliably store large content that isn’t really critical. As you’ve noted, media is one good example of this. In my case, I’m using a RAIDZ1 vdev of 3 disks, with the understanding that things could go south during a resilver.
- nix23 6y ago
  Yeah true, in my Home-server i have one mirror (2 disks for OS/Swap etc) and one RaidZ1 (5 Disks for data), for me it's just important that i have no bit-rot etc. Daily backups (restic) are made so no problem whatsoever.
nix23 6y ago
I then tell a real horror-story about my past, not so long ago there was a EVA (A SAN from HP) with a freaking out hardware Raid out of nowhere (uptime about 200 days no Firmware update before nothing) it just started to freak out, the story goes on but the lesson...No more HW-Raid/proprietary stuff for me EVER again.
- labawi 6y ago
  While none for me either, I used to look down on HW RAID etc. Lately I've realized, such solutions are genuinely useful and more likely to survive/work properly, if the people managing it do not understand the technology, don't want to understand or just don't have the time for it. HW RAID with red blinking lights seems more likely to survive than a SW RAID that works until it doesn't, because you lost all the redundancy and noone there knows how either work.
  - nix23 6y ago
    >HW RAID with red blinking lights seems more likely to survive than a SW RAID that works until it doesn't, because you lost all the redundancy and noone there knows how either work. Well if one ever sees your red blinking light because the new printer stands in front of the server :)
seized 6y ago
People act like resilvers on parity are vastly different than on mirrored. They are not, its just reads and writes, granted in a different pattern. My RAIDZ2 resilvers at about what I expect the old drives in it to write at, ~100MB/sec. In a mirrored setup it all depends on the one drive doing the reading to survive the process. The drive that is most likely the same age as the one that just failed. A resilver isnt more likely to kill an existing drive than a scrub. And the general advice is to do those semi regularly. A blanket statement equating parity RAID to "more risky" is absurd and leads to people thinking they need a massive pool of mirrors just for their home files. ZFS has numerous options to prioritize different IO higher or lower (priority to resilvers or end user data).

Cyph0n 6y ago

Critical data for the average user typically does not require much space, so the cost of storage overhead in a mirrored ZFS pool is not too bad. RAIDZ is much more cost efficient when you need to reliably store large content that isn’t really critical. As you’ve noted, media is one good example of this. In my case, I’m using a RAIDZ1 vdev of 3 disks, with the understanding that things could go south during a resilver.

nix23 6y ago
Yeah true, in my Home-server i have one mirror (2 disks for OS/Swap etc) and one RaidZ1 (5 Disks for data), for me it's just important that i have no bit-rot etc. Daily backups (restic) are made so no problem whatsoever.

nix23 6y ago

I then tell a real horror-story about my past, not so long ago there was a EVA (A SAN from HP) with a freaking out hardware Raid out of nowhere (uptime about 200 days no Firmware update before nothing) it just started to freak out, the story goes on but the lesson...No more HW-Raid/proprietary stuff for me EVER again.

labawi 6y ago
While none for me either, I used to look down on HW RAID etc. Lately I've realized, such solutions are genuinely useful and more likely to survive/work properly, if the people managing it do not understand the technology, don't want to understand or just don't have the time for it. HW RAID with red blinking lights seems more likely to survive than a SW RAID that works until it doesn't, because you lost all the redundancy and noone there knows how either work.
- nix23 6y ago
  >HW RAID with red blinking lights seems more likely to survive than a SW RAID that works until it doesn't, because you lost all the redundancy and noone there knows how either work. Well if one ever sees your red blinking light because the new printer stands in front of the server :)

seized 6y ago

People act like resilvers on parity are vastly different than on mirrored. They are not, its just reads and writes, granted in a different pattern. My RAIDZ2 resilvers at about what I expect the old drives in it to write at, ~100MB/sec. In a mirrored setup it all depends on the one drive doing the reading to survive the process. The drive that is most likely the same age as the one that just failed. A resilver isnt more likely to kill an existing drive than a scrub. And the general advice is to do those semi regularly. A blanket statement equating parity RAID to "more risky" is absurd and leads to people thinking they need a massive pool of mirrors just for their home files. ZFS has numerous options to prioritize different IO higher or lower (priority to resilvers or end user data).