FreeNAS Without the RAID

[danb35]

rapped by the dunning-Kruger effect.

The problem with citing the Dunning-Kruger effect is that the people who are demonstrating it don't understand they're doing so--exactly as the DKE would predict.

 

[Evertb1]

The problem with citing the Dunning-Kruger effect is that the people who are demonstrating it don't understand they're doing so--exactly as the DKE would predict.

The real problem with citing the Dunning-Kruger effect is that it is a subject belonging in the field of psychology. Now personaly I think the OP is a bit stubbern (to say the least). But I respect this forum because of the knowledge of its members about FreeNAS and related subjects and not about psychology. Should we not just leave it at that and come to the conclusion that only those that want to be helped can be helped?

 

[garm]

only those that want to be helped can be helped?

I guess that is a better statement :)

 

[carlmart]

But as I said, I also have the guidance of a fully professional production company person, and his views seem to be different from yours.

He was the one that recommended using a desktop as server, with FreeNAS as the program.

 

[carlmart]

I am stubborn, the same as you are, when I have to be.

About the help, you can choose if the help proposed is really the only one acceptable or possible, because this doesn't seem to be the case here.

 

[HoneyBadger]

Well this thread has certainly gone down an interesting little tangent.

Summarized simply, "we'll never reach an agreement" here because you think data loss is acceptable.

 

[garm]

He was the one that recommended using a desktop as server, with FreeNAS as the program.

I would love if this person could start a let’s build thread, I guess some of the “real” sysadmins (and the hobbyist like myself) would have some comments ;)
Who’s authority you recognize is up to you, in the end reality is judge and executioner.

 

[Yorick]

I think the answer has been given. OP wants to lose the data on one disk should that disk fail. And then restore from Blu-ray / dvd backup.

FreeNAS can do that. It wasn’t designed to work this way, but it will. Solution:
Single-drive vdev, each vdev in its own pool.
Result: Failure of a single drive will wipe out the contents of just that drive.
Unlikely recovery scenario: Drive indicates it will fail, add a mirror drive to the vdev and let it resilver. In the exceedingly unlikely event that this succeeds before the drive fails, remove the drive that was about to fail from the mirror and run on a single drive vdev again
Necessary precaution: Try this procedure once with just a small test file on a functioning drive; write down the steps you needed to take so you can do this without needing to learn how to in the case of a case.
Most likely outcome in case of disk failure: Loss of the contents of that one disk.

Alternative scenarios.

Raidz1, one redundancy drive. Likelihood of loss of all data over 10 years roughly 4.4% if assuming manufacturers lie and the drive MtBF should be assumed to be 250k; fraction of a percent if 2.5m is the real number.

Raidz2, two redundancy drives. Likelihood of loss of all data over 10 years at 1 percent or below. May be perfectly acceptable risk particularly with a backup in place.

Drives to use: Shuck some WD elements, there are HGST He10 enterprise drives inside. Can’t get better for the money.

 

[Chris Moore]

I also have the guidance of a fully professional production company person

You discount the voice of experience from many people in favor of one person because they are "professional" without realizing that many of the members here are also working professionals.

I think this thread is not contributing anything useful and needs to be closed.

I am stubborn, the same as you are, when I have to be.

You don't need to be stubborn, but you choose to be.