'root filesystem is not BTRFS. During Rockstor installation, you must select BTRFS instead of LVM and other options for root filesystem. Please re-install Rockstor properly.'

Support
1

[Please complete the below template with details of the problem reported on your Web-UI. Be as detailed as possible. Community members, including developers, shall try and help. Thanks for your time in reporting this issue! We recommend purchasing commercial support for expedited support directly from the developers.]

Brief description of the problem

New install, 10 2TiB HDDS, perc 5/i IT mode SAS controler

Detailed step by step instructions to reproduce the problem

Accessed GUI, setup user, got the error

Web-UI screenshot

[Drag and drop the image here]

Error Traceback provided on the Web-UI

rockstor_error.png1427×933 146 KB
2

Same problem?

Install Rockstor 3.8.15-0 to USB Stick 32GB with UEFI.
Shutdown -> add 1TB SATA HDD x 4 to motherboard. (LEDed)
Boot with Rockstor normaly.
logon -> storage -> Disks "No disks added. Click on Rescan to discover disks"
Rescan -> Houston

Upgrade to Rockstor 3.8.15-2 -> same problem (now here)

        Traceback (most recent call last):

File “/opt/rockstor/src/rockstor/rest_framework_custom/generic_view.py”, line 40, in _handle_exception
yield
File “/opt/rockstor/src/rockstor/storageadmin/views/disk.py”, line 272, in post
return self._update_disk_state()
File “/opt/rockstor/eggs/Django-1.6.11-py2.7.egg/django/db/transaction.py”, line 371, in inner
return func(*args, **kwargs)
File “/opt/rockstor/src/rockstor/storageadmin/views/disk.py”, line 62, in _update_disk_state
disks = scan_disks(settings.MIN_DISK_SIZE)
File “/opt/rockstor/src/rockstor/system/osi.py”, line 120, in scan_disks
base_root_disk = root_disk()
File “/opt/rockstor/src/rockstor/system/osi.py”, line 768, in root_disk
raise NonBTRFSRootException(msg)
NonBTRFSRootException: ‘root filesystem is not BTRFS. During Rockstor installation, you must select BTRFS instead of LVM and other options for root filesystem. Please re-install Rockstor properly.’

PS. In the root console on Rockstor, lsblk shows added HDDs as /dev/sda-sdd. (/sde is USB stick)

3

@m3elloa Hello again and a very belated Welcome to the Rockstor community.

ZX80 in you profile image, fancy. Not sure but I think my coding days began on a ZX81, basic and machine code. Quite a mix.

To your issue: This message is pretty much what it states, the system appears to believe that it’s root ‘/’ is mounted on a non btrfs filesystem. This is a more recent enforcement and allows us to supports snapshots prior to upgrades so that we might do upgrade roll backs. I believe @sfranzen is working on the foundations of this system (in issue #1432) with his integration of snapper into Rockstor, quite exciting.

Anyway if you could confirm a non btrfs root by pasting the following command output we can see if we have a bug here or not:

lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL

Thanks for reporting.

4

@yukimi Welcome to the Rockstor community, and thanks for reporting.

My reply in this thread to @m3elloa should also relate to your install, although I’m not sure of the UEFI install side of Rockstor.

Could you also paste the output of that same command and we can see if this is a bug. As we essentially use a re-badged and slightly older installer from CentOS7 we don’t fully take account of all it’s install variation. So it’s probably best to go with the defaults as they are pre-defined and understood by the subsequent Rockstor system.
ie the Rockstor drive is expected to be installed as follows, using an excerpt from the same command:

lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL

NAME   SERIAL   SIZE TRAN   TYPE  FSTYPE            LABEL
sda    QM00005    8G sata   disk                    
├─sda1          500M        part  ext4              
├─sda2          820M        part  swap              
└─sda3          6.7G        part  btrfs             rockstor_rockstor

ie in non UEFI and default partitioning (non LVM), in this case on an 8GB drive.
where /boot is the sda1 partition:

mount | grep 'boot'
/dev/sda1 on /boot type ext4 (rw,relatime,data=ordered)

Thanks.

5

@phillxnet, thanks for your reply.

Yes, around since z-80 synclair and TRS, motorola 65000 Apple II … good times coding in assembly and saving on k-7, but back to the point :slight_smile:

I was just doing a fresh install to move my Rockstor from my VM to a physical box for a production setting. I didn’t change anything on the partitions, so whatever it was created by the install is what I have. Before the install I did download the latest version used for it.

I’ll reinstall and let you know what I’ll get from that output, but again, no changes were made on the partitioning scheme.

Best regards and stand by for more :slight_smile:

6

Sorry guys. Will take longer than expected as I have to play with the bios configuration and change some UEFI setting. Now even anaconda is giving me trouble. As soon I get it going I’ll reinstall and retest.

7

There you go.

After re-install and let the installer partition the pool I got the same error and here’s the output for lsblk as requested.

[root@Rockstor ~]# lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL
NAME SERIAL SIZE TRAN TYPE FSTYPE LABEL
sda JK1105YAKYG9TX 1.8T sata disk
├─sda1 500M part ext4
└─sda2 1.8T part LVM2_member
├─rockstor_rockstor-swap
11.8G lvm swap
├─rockstor_rockstor-root
50G lvm xfs
└─rockstor_rockstor-home
18.1T lvm xfs
sdb YAK71UHV 1.8T sata disk
└─sdb1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdc JK11B1B9HRP24F 1.8T sata disk
└─sdc1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdd JK11B1B9K3ENRF 1.8T sata disk
└─sdd1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sde Y485REZTS 1.8T sata disk
└─sde1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdf YAK6D79V 1.8T sas disk
└─sdf1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdg JK11B1YBKYBTHF 1.8T sas disk
└─sdg1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdh JK1105B8G6Z14X 1.8T sas disk
└─sdh1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdi WD-WCAVY4457087 1.8T sas disk
└─sdi1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sdj JK1171YAJSMAKS 1.8T sas disk
└─sdj1 1.8T part LVM2_member
└─rockstor_rockstor-home
18.1T lvm xfs
sr0 K1HEAEA5206 729M sata rom iso9660 Rockstor 3 x86_64

8

2nd try, installing on one HD only, got a new issue: No IP on CLI. Not sure if regression of 924 …

Interface get the assigned IP from my router (pfSense):

enp6s0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 10.10.10.18 netmask 255.255.255.0 broadcast 10.10.10.255
inet6 fe80::215:17ff:fedc:2c5e prefixlen 64 scopeid 0x20
ether 00:15:17:dc:2c:5e txqueuelen 1000 (Ethernet)
RX packets 17587 bytes 24680606 (23.5 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 8570 bytes 627883 (613.1 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 18 memory 0xf7a20000-f7a40000

lsblk output below (installed only on sda, but haven’t deleted old partitions, so still listed there - sorry)

[root@Rockstor ~]# lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL
NAME SERIAL SIZE TRAN TYPE FSTYPE LABEL
sda JK1105YAKYG9TX 1.8T sata disk
├─sda1 500M part ext4
├─sda2 11.8G part swap
└─sda3 1.8T part btrfs rockstor_rockstor00
sdb YAK71UHV 1.8T sata disk
└─sdb1 1.8T part btrfs rockstor_rockstor
sdc JK11B1B9HRP24F 1.8T sata disk
└─sdc1 1.8T part btrfs rockstor_rockstor
sdd JK11B1B9K3ENRF 1.8T sata disk
└─sdd1 1.8T part btrfs rockstor_rockstor
sde Y485REZTS 1.8T sata disk
└─sde1 1.8T part btrfs rockstor_rockstor
sdf YAK6D79V 1.8T sas disk
└─sdf1 1.8T part btrfs rockstor_rockstor
sdg JK11B1YBKYBTHF 1.8T sas disk
└─sdg1 1.8T part btrfs rockstor_rockstor
sdh JK1105B8G6Z14X 1.8T sas disk
└─sdh1 1.8T part btrfs rockstor_rockstor
sdi WD-WCAVY4457087 1.8T sas disk
└─sdi1 1.8T part btrfs rockstor_rockstor
sdj JK1171YAJSMAKS 1.8T sas disk
└─sdj1 1.8T part btrfs rockstor_rockstor
sr0 K1HEAEA5206 729M sata rom iso9660 Rockstor 3 x86_64
[root@Rockstor ~]#

What steps you guys want me to try?

9

After re-install on one HDD, the problem went away.

Is this a limitation on the number of disks (10), pool size (20 TiB)?

What can I do to add the other 9 HDDs to the pool?

10

@m3elloa Hello again. My you’ve been busy.

OK, sorry I didn’t notice / ask if you were installing to multiple devices during the installer for the system drive.
As is our re-badged anaconda installer defaults to using LVM and XFS when setting up multiple devices so that explains things nicely in your instance. We do have an advanced user how-to over at the official docs entitled: Mirroring Rockstor OS using Linux Raid which involves twisting the installers arm a little and using the recovery system in order to wangle a mirrored install arrangement. Bit of a pain and a long way from elegant but may serve your requirements in that area.

So as is your partitions on last report look dandy except that your btrfs data drives are partitioned.
Rockstor deals purely with full disk btrfs as this keeps things simple and allows us to discard the mess that is decades of partitioning systems, at least on the btrfs data disks side.

Not as far as I’m aware but in reply to:

I’m not sure of your arrangement here, I’m assuming that the 10 disks were partitioned by the installer and are essentially empty in which case you need to use the wipefs feature built into the little cog icon on the Disk page (or manually if you prefer) so that the drives are returned to a usable state (ie blank): given the above caveat of only full disk btrfs. You should then be able to create a btrfs pool from the now usable disks.

Obviously if these disks have data on then we are in a very different position.

Let us know what the situation is and how you get on.

Hope that helps.

Also the Disks and Pools sections of the docs may well help.

11

Thanks for replying.

Not as busy as coding on z80, but some :slight_smile:

Nicely explained, thanks again. I have it running installed on a single drive on the pool and created a pool with the other ones using the unofficial Raid6 (You would say: Are you crazy… well it’s a test box with data from the other two servers in rsync).

I’m sure that it’s explained in the docs and I should have read it, but is there a way to avoid this for regular users? Not allow the install to be set on multiple HDDs, I mean.

Now that I understand, I’m going to re-do all using a SSD to be faster and let the HDDs to be used for the pool.

Thanks again!

12

@m3elloa Hello again and thanks, glad your up and running.

I don’t know but so that it might be put up for consideration at least I’ve opened the following issue in the rockstor-iso repo:

This may tie in with the recent:


as it turns out that the current iso is based on an older version of the CentOS install iso.

Hope that helps.

13

I was having the same issue

First thing I realized is I had installed from a USB created with RUFUS. Though the install would run fine, I was having issues with cleaning up partitions. I recall there being some nuance to USB installs and sure enough, the Quick Start states:

Please note the following USB image writing programs have been found to produce NON working USB install disks when used with the Rockstor iso and their default settings.

Unetbootin
Rufus (N.B. does work in DD image mode, see our Rockstor USB install disk using Rufus)

I also used gparted to wipe any remaining partitions on the SSD and reinstalled from a USB created with Rawrite32.and BAM! No more “‘root filesystem is not BTRFS” messages.

1 Like
14

Sorry for late to reply because this is my weekend private mission in the lab.
I found after wrote previous reply that boot USB volume is corrupted with old EFI parted SATA HDD.
(OS partition has /sde is not right status)

Today I did fresh install to USB stick again. But Rockstor do not display IP address on the top now.
Its taking more time by this regression.

Yukimi

15

@yukimi

That a funny one, did you do any network config within the installer, if so you could try a re-install and leave the default dhcp while installing as Rockstor has it’s own network config once installed. Might be a red herring but worth checking out.

Edit: I’ve just noticed the same here, but not sure when this regression occurred. Do you have an info on the Rockstor version when the ip did appear in the top bar, and what leads to you believe this is the cause of a slow down.

@Flyer I know you are neck deep but does this ‘no-ip in header bar’ ring any bells for your, also I don’t think we have an issue open on this.

Thanks.

16

Hi @yukimi & @phillxnet, if I’m not wrong that no-ip in header bar started with net views updates:
setting the management ip address should solve it

M.

1 Like
17

Thanks for your comment.
Today I reinstall Rockstor 3.8.15 again and I found some of result.

(1) Boot first time before logon into Rockstor host via browser, IP does not show top of the host.
(2) Boot second time after logon into Rockstor host via browser, IP does show top of the host.
*if my installed OS is not corrupted :slight_smile:

Anyway my facing problem is out of this topic. Please ignore this. This is okay.

back to the topic… My lsblk is now here.

[root@rockstor ~]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 931.5G 0 disk
└─sda1 8:1 0 931.5G 0 part
sdb 8:16 0 931.5G 0 disk
└─sdb1 8:17 0 931.5G 0 part
sdc 8:32 0 931.5G 0 disk
└─sdc1 8:33 0 931.5G 0 part
sdd 8:48 0 931.5G 0 disk
└─sdd1 8:49 0 931.5G 0 part
sde 8:64 1 29.3G 0 disk
├─sde1 8:65 1 200M 0 part /boot/efi
├─sde2 8:66 1 500M 0 part /boot
└─sde3 8:67 1 28.7G 0 part
├─rockstor_rockstor-swap 253:0 0 3G 0 lvm [SWAP]
└─rockstor_rockstor-root 253:1 0 25.7G 0 lvm /

18

Hi @yukimi,
talking about that no-IP, I’m quite sure about the management interface setting, but @suman had the new networking code so asking him if can confirm :slight_smile:

19

Thanks for reply. Now I am ok because I can access Web-UI in this time.

P.S.
‘root filesystem…’ problem is not solved. Is it need to re-install with all of disks are online for to be BTRFS?

1 Like
20

@yukimi Hello again.
If I understand correctly you are still getting the root filesystem is not BTRFS message.
In you lsblk command you are missing the FSTYPE option. This would be helpful ie:

lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL,MOUNTPOINT

Also if you put a tripple grave accent character “`” on the line before and after your output when you past it into this forum your output will be easier to read as it should then appear like the following:

lsblk -o NAME,SERIAL,SIZE,TRAN,TYPE,FSTYPE,LABEL,MOUNTPOINT
NAME      SERIAL                SIZE TRAN   TYPE  FSTYPE            LABEL             MOUNTPOINT
sda       sys-drive-serial-num    8G sata   disk                                      
├─sda1                          500M        part  ext4                                /boot
├─sda2                          820M        part  swap                                [SWAP]
└─sda3                          6.7G        part  btrfs             rockstor_rockstor /mnt2/rockstor_rockstor

Note there are no LVM entries and that the root partition (sda3) is on btrfs and no uefi in the above example.

Yes it is necessary to re-install if this is not the case, also all data drives can be disconnected during install and one can then re-connect after successful install and initial setup. That way there is no risk to existing data.

Best to go with defaults during partitioning and to wipe all existing partitions first such as is indicated in the official docs in the Reinstalling Rockstor section. Hence the advise to remove / disconnect data drives first; just in case you wipe the wrong drive during install / re-install. It’s best to read the whole of this section first as it does need a little re-ordering which should come in time.

In the Installation section of the docs there is also:

“See our YouTube VirtualBox Rockstor install demo or our Rockstor in Virtual Machine Manager guide”

I’ve just noticed that the first video link is broken, probably because it’s really out of date now and I’ve opened an issue for this:


EDIT: the above issue has now been closed.

Hope that helps.