How is the disk being used?

[AAronoff]

FOA, I'm new to FreeNAS and I'm working with an installation that I inherited. It's using v9.2.1.5. (I'll upgrade it later.)

Under Volumes | View Disks, there are three disks installed: two 1 TB Toshiba drives (ada0 and ada1) and one 2 TB Western Digital (ada2).

Under Storage | View Volumes, there is a single volume (A1) with three datasets (followed by "jails"). When I click on the volume name and then the "Volume Status" button, A1 shows "stripe". I presume that's the two Toshiba drives in RAID 0.

How can I confirm that the striped set is, in fact, the two Toshiba drives?

How can I find out how the 2 TB WD disk is being used?

Is there any way to see which datasets are stored on the drives?

TIA for your help.

 

[danb35]

Can you post either a screenshot of the Volume Status screen, or the output of zpool status in code tags?

 

[AAronoff]

FOA, thanks for your reply.

Here's the output of zpool status (I'm not sure what "in code tags" means, so I ignored it):

Code:

zpool status
  pool: A1
state: ONLINE
  scan: scrub repaired 0 in 6h25m with 0 errors on Sun Jun 25 06:25:32 2017
config:
   NAME  STATE  READ WRITE CKSUM
  A1  ONLINE  0  0  0
  gptid/f3275cc1-d1a3-11e3-b9e1-001e8c833578  ONLINE  0  0  0

errors: No known data errors

HTH.

regards, Andy

 

[danb35]

"In code tags" means that you click the "code" button in the little toolbar above where you type your reply. It inserts tags like this: [ CODE][ /CODE] (but without the spaces). Anything you type between them will have its formatting preserved, making things like zpool status and SMART output much easier to read. But in your case, it's not much of an issue--you only have one disk in the pool. To find out which one it is, the volume status page will give you the disk ID (something like adanp2, probably). From that, you can look at the View Disks screen to see which disk adan is.

 

[danb35]

And as to your question about datasets, all datasets within a pool are stored on that pool, but there's no distinction among the members of the pool--so you wouldn't have one dataset on one disk, and two on another.

 

[gpsguy]

Please proved detailed hardware information per the forum rules.

Are you running FreeNAS with hardware RAID?

 

[AAronoff]

danb35 – thanks for the replies.

Volume Status shows A1 | stripe | ada2p2

View Disks shows that ada2 is the 2 TB Western Digital.

Thus, the two Toshiba disks appear to be unused.

Why does the Volume Status indicate "stripe" for a single disk?

all datasets within a pool are stored on that pool, but there's no distinction among the members of the pool

Do I understand correctly that a pool is dedicated to a particular disk (or RAID set)?

gpsguy – As for hardware RAID, if a single disk is being used out of the three, the question is, I believe, moot.

regards, Andy

 

[danb35]

Why does the Volume Status indicate "stripe" for a single disk?

Because that's pretty much the default for ZFS. If it isn't a mirror, or a RAIDZn, it's a stripe.

Thus, the two striped Toshiba disks appear to be unused.

Fixed that for you--as they appear to be unused, they aren't striped.

Do I understand correctly that a pool is dedicated to a particular disk (or RAID set)?

Not exactly. A pool consists of one or more vdevs. A vdev consists of one or more disks (or other devices, but let's don't go there yet). If a vdev has more than one disk, it's in some sort of redundant configuration--either mirrored or RAIDZn. The pool is striped across all vdevs. You can add vdevs to a pool at any time, but once a vdev has been added, it can never be removed.

 

[AAronoff]

danb35 — Again, thanks for your help.

So, we have three disks, two of which are idle.

Is there any reason not to shut down the server and remove the idle disks?

The single used disk is labeled "stripe", but, in fact, it's not. Obviously, all the datasets are stored on the single drive.

That drive is reported to be over 80% full and FreeNAS is issuing an alert. What's the best way to increase capacity? Add a 2 TB disk and stripe it? (9.2.1 User's Guide, page 121) Is there a tutorial here or a forum post or a YouTube video you can recommend?

TIA

regards, Andy

 

[danb35]

Is there any reason not to shut down the server and remove the idle disks?

Not really, unless...

What's the best way to increase capacity? Add a 2 TB disk and stripe it?

...you could always stripe in the two idle disks instead, if you don't care too much about your data. Of course, a striped pool means that if any of your disks fails, you lose everything on your pool. But if you're OK with that, that's probably the simplest way to expand your pool.

 

[AAronoff]

you could always stripe in the two idle disks instead

I strongly suspect, based on sparse documentation, that the two Toshibas were taken out of service due to failure. IAC, it's not worth wasting any time testing them. The need for mirroring is abated via cloud backup of critical files.

Any recommendations for learning how to add another disk?

Once the pool is extended to 4 TB, I presume mirroring could be provided by adding a single 4 TB disk (or two 2 TB disks). Is this the way it works?

regards, Andy

 

[danb35]

Any recommendations for learning how to add another disk?

I'm pretty sure it's covered in the manual, but go to the volume manager, choose your pool under the Volume to Extend drop-down, and add the disk.

 

[AAronoff]

I'm pretty sure it's covered in the manual

I didn't find much more than a couple of sentences on page 121 of the 9.2.1 User's Guide. Just checking to make sure no nasty surprises lie in wait.

Thanks again for your help!

regards, Andy

 

[Stux]

The need for mirroring is abated via cloud backup of critical files.

So, *when* the drive fails and you lose the lot, how long is it going to take you to rebuild and reconfigure and restore the server...

The best thing to do is to buy two 4TB drives. Add one as a mirror to the 2TB drive, then replace the 2TB drive with the other.

You'll end up with double the space and infinitely more redundancy.

 

[AAronoff]

When one of the disks fails, essential data will be immediately available over the web. The client will continue to run their business. This server hasn't been managed for a few years and it needs work. When FreeNAS is finally configured correctly, I'll try to convince the customer to go to RAID 1. First, we need to set up users and groups to reflect the current working environment and add storage so that files can be put somewhere without throwing alerts. (Sorta basic stuff for a file server. ;) )

regards, Andy

 

[Arwen]

Hey guys, the original poster is using an old version of FreeNAS, which still supports UFS.

Perhaps the 2 disks are used by UFS volumes?

 

[AAronoff]

Hi, Arwen. How would I check that?

regards, Andy

 

[Arwen]

@AAronoff
Sorry, I've never used FreeNAS 9.2.x or FreeNAS with UFS.

However, if you supply the output of df -h in code tags, we may get a clue. Meaning it is very likely your ZFS volume is sitting on /mnt/A1/..., so if you have something else mounted at /mnt/X that has a maximum capacity of 1TB, that would indicate a UFS Mirror.

Plus, the manual mumbles something about a UFS Volume manager. I did not look at it closely, sorry.

 

[AAronoff]

The "df -h" ouput may be scrolling. If I direct it to a file with ">" , I'm not sure where the file will be placed. (OS ignorance shows up at the first console command. :( )

Code:

system/samba4
A1/.system/syslog  114G  8.7M  114G  0%  /mnt/A1/.system/syslog
A1/Esperso  114G  164k  114G  0%  /mnt/A1/Esperso
A1/MTimeM  1.4T  1.3T  114G  92%  /mnt/A1/MTimeM
A1/Win  477G  163G  314G  34%  /mnt/A1/Win
A1/jails  114G  404k  114G  0%  /mnt/A1/jails
A1/jails/.warden-template-pluginjail  115G  702M  114G  1%  /mnt/A1/jails/.warden-template-pluginjail
A1/jails/bacula-sd_1  115G  883M  114G  1%  /mnt/A1/jails/bacula-sd_1
A1/jails/crashplan_1  115G  1.0G  114G  1%  /mnt/A1/jails/crashplan_1
A1/jails/plgt  115G  952M  114G  1%  /mnt/A1/jails/plgt
devfs  1.0k  1.0k  0B  100%  /mnt/A1/jails/plgt/dev
procfs  4.0k  4.0k  0B  100%  /mnt/A1/jails/plgt/proc

Thankfully, this is not a complicated system.

TIA for your help.

regards, Andy

 

[danb35]

That doesn't look like there are any other volumes--everything appears to be in A1, which is your ZFS pool.