ZFS trouble shooting issues for newbies?

General Discussion
1

Other than referring new users with problems to guides that show how to make an effective request for help, any other ideas?

We are still seeing several people a week very new to ZFS, that seem to panic when they can’t get a fast answer. They then start doing bad things, like remaking their pools, (either knowing they need to restore afterwards, OR NOT!).

In my opinion, all TrueNAS users should at one point either know or read the ZFS primer and terminology resources.

I think some of these are at issue:

  • ZFS pool import is so different from normal file system mounting, some new users don’t understand it.
  • They had a power loss, and assume it trashed their pool. Except ZFS was specifically designed to NOT do that.
  • Using VM, (especially ProxMox), without a clear understanding of how to make a data safe VM that uses ZFS. (Pass through the disk controller and blacklist the controller to the hypervisor).
  • ZFS Dataset encryption is different from some other types, (full disk, file based). Of course, the community here does not have as much experience with it, since not all of us use ZFS encryption.
  • Use of poorly selected hardware, (like a NUC with USB attached data drives).

Perhaps HexOS will help these users.

I am not trying to be an elitist, but ZFS is hard core server software, Even as an experience Unix SysAdmin, it can be hard to understand. At some jobs I have had, new Unix SysAdmins tend to have Linux experience, but not AIX or Solaris, because those basically require a software license to use. (And AIX requires special hardware too.)

For example, I’ve had to give multiple talks and have several teaching guides for new Unix SysAdmins at my day job about Solaris & ZFS. These new SAs tend to think Linux is the best thing since sliced bread, up until there are problems. Solaris 11’s alternate boot environments have solved multiple problems, but creates others. Like trying to remove huge log files from “/var” without understanding the ABE’s snapshot is holding onto the data.

So, any other problems you see?
And more importantly, any solutions?

Our request for information is a good start. As are the various resources.

Perhaps we need to see if this new forum software supports a link below the “Forum Rules”, so a “Reporting a Problem” link could be added.

2

There’s no way around human nature. I have come to accept that there’s going to be a lot of causalities and collateral damage, which cannot be avoided.

Here’s the problem:

Q: “How do we get users, who are new to ZFS, to avoid destructive accidents and better understand how to manage a ZFS-based NAS server?”

A: “Make a guide that covers the common issues!”

Q: “How do we get them to read the guide before committing?”

A: “…”

:wink:

1 Like
3

Good point.

Some make the problems, (unknowingly or knowingly), before they come here to the forums. Then start asking for help. Even a few making it harder by saying “help” without enough details.

We have just seen too many newbies hop on to TrueNAS SCALE “because Linux!!!”. But, other than newer development happening on SCALE, it is still a NAS appliance first. Not a general purpose Linux distro.

Your probably right. Their is low chance of preventing problems for most new users, (that end up having problems). But I am trying by asking the question!