Partitioning :: Automatic Mode :: Antergos installation on ZFS failed

[satriani-vai]

"Error running ['zfs', 'create', '-V', '-7072G', '-b', '4096', '-o', 'primarycache=metadata', '-o', 'checksum=off', '-o', 'com.sun:auto-snapshot=false', 'antergos_wzpp/home']: bad volume size '-7072G': bad numeric value '-7072G'"

EDIT: another one Cannot wipe the filesystem of device /dev/sda: wipefs: error: /dev/sda: probing initialization failed: Device or resource busy

[karasu]

Obviously, first one is a bad calculation. Can you tell me how big is your harddrive?

Second one is because you didn't reboot your computer before retrying, right?

[satriani-vai]

I have a 1TB hybrid harddrive

[lots0logs]

@karasu I don't have time to do it now so if you can get to it before me then feel free... Let's create this file after we've finished formatting/partitioning: /var/tmp/.cnchi_partitioning_completed. We can check for the existence of the file in check.py and prevent installation from continuing (eg. require a reboot). We dont need to add a StateBox for it though. Lets just popup a message dialogue. I think that would be easier so we dont have to deal with GTK Window Size issues.

[satriani-vai]

this issue is still outstanding http://i.imgur.com/KLQB1va.png

[lots0logs]

@satriani-vai I performed two successful ZFS installations yesterday. Could you provide the logs for your failed attempts please? Also, just to be clear, only the latest testing ISO images will work for ZFS.

[satriani-vai]

Of course the last testing ISO should do the trick. Could you please send me a link to the ISO please?

[lots0logs]

Sure! http://build.antergos.com/browse/testing

[satriani-vai]

ISO won't boot from USB :( I will try tomorrow's build.

[satriani-vai]

OK I've got it with another USB stick, but this issue is still outstanding with latest testing ISO Here the output:

...
2016-02-08 13:18:31 [INFO] pac.py(530) cb_log(): adding 'fakeroot' to package cache for db 'local'
2016-02-08 13:18:31 [DEBUG] pac.py(528) cb_log(): added 837 packages to package cache for db 'local'
2016-02-08 13:18:31 [DEBUG] pac.py(528) cb_log(): loading group cache for repository 'community'
2016-02-08 13:18:48 [DEBUG] zfs.py(639) run_format(): Configuring ZFS in /dev/sda
2016-02-08 13:18:48 [DEBUG] zfs.py(585) init_device(): Zapping device /dev/sda...
2016-02-08 13:18:49 [INFO] slides.py(242) manage_events_from_cb_queue(): 
2016-02-08 13:18:49 [INFO] slides.py(242) manage_events_from_cb_queue(): Getting your disk(s) ready for Antergos...
2016-02-08 13:18:50 [DEBUG] run_cmd.py(70) call(): 1953525168
2016-02-08 13:18:51 [DEBUG] run_cmd.py(70) call(): no pools available to import
2016-02-08 13:18:51 [DEBUG] run_cmd.py(70) call(): no pools available to import
2016-02-08 13:18:56 [DEBUG] zfs.py(1000) create_zfs(): Cnchi will create a ZFS pool using /dev/sda3 devices
2016-02-08 13:18:56 [DEBUG] zfs.py(933) create_zfs_pool(): Creating zfs pool antergos_hom2 of type none
2016-02-08 13:18:56 [DEBUG] zfs.py(957) create_zfs_pool(): Pool antergos_hom2 created.
2016-02-08 13:18:56 [WARNING] zfs.py(869) set_zfs_mountpoint(): cannot mount '/': directory is not empty
property may be set but unable to remount filesystem
2016-02-08 13:18:56 [DEBUG] zfs.py(778) get_pool_size(): zpool list -H -o size antergos_hom2
2016-02-08 13:18:56 [DEBUG] zfs.py(1042) create_zfs(): Creating zfs subvolume 'home' (899GB)
2016-02-08 13:18:56 [DEBUG] zfs.py(850) create_zfs_vol(): Creating a zfs vol antergos_hom2/home of size 899GB
2016-02-08 13:18:56 [ERROR] run_cmd.py(85) call(): Error running ['zfs', 'create', '-V', '899G', '-b', '4096', '-o', 'primarycache=metadata', '-o', 'checksum=off', '-o', 'com.sun:auto-snapshot=false', 'antergos_hom2/home']: cannot create 'antergos_hom2/home': out of space
2016-02-08 13:18:57 [ERROR] process.py(129) run(): "Error running ['zfs', 'create', '-V', '899G', '-b', '4096', '-o', 'primarycache=metadata', '-o', 'checksum=off', '-o', 'com.sun:auto-snapshot=false', 'antergos_hom2/home']: cannot create 'antergos_hom2/home': out of space"
2016-02-08 13:18:57 [ERROR] process.py(133) run(): Traceback (most recent call last):
2016-02-08 13:18:57 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/misc/run_cmd.py", line 66, in call
    timeout=timeout)
2016-02-08 13:18:57 [ERROR] process.py(133) run():   File "/usr/lib/python3.5/subprocess.py", line 629, in check_output
    **kwargs).stdout
2016-02-08 13:18:57 [ERROR] process.py(133) run():   File "/usr/lib/python3.5/subprocess.py", line 711, in run
    output=stdout, stderr=stderr)
2016-02-08 13:18:57 [ERROR] process.py(133) run(): subprocess.CalledProcessError: Command '['zfs', 'create', '-V', '899G', '-b', '4096', '-o', 'primarycache=metadata', '-o', 'checksum=off', '-o', 'com.sun:auto-snapshot=false', 'antergos_hom2/home']' returned non-zero exit status 1
2016-02-08 13:18:57 [ERROR] process.py(133) run(): 
During handling of the above exception, another exception occurred:
2016-02-08 13:18:58 [ERROR] process.py(133) run(): Traceback (most recent call last):
2016-02-08 13:18:58 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/installation/process.py", line 98, in run
    self.install_screen.run_format()
2016-02-08 13:18:58 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/installation/zfs.py", line 746, in run_format
    self.create_zfs(solaris_partition_number)
2016-02-08 13:18:58 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/installation/zfs.py", line 1043, in create_zfs
    self.create_zfs_vol(pool_name, "home", home_size)
2016-02-08 13:18:58 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/installation/zfs.py", line 859, in create_zfs_vol
    call(cmd, fatal=True)
2016-02-08 13:18:58 [ERROR] process.py(133) run():   File "/home/antergos/Downloads/Cnchi/cnchi/misc/run_cmd.py", line 87, in call
    raise InstallError(msg)
2016-02-08 13:18:58 [ERROR] process.py(133) run(): misc.extra.InstallError: "Error running ['zfs', 'create', '-V', '899G', '-b', '4096', '-o', 'primarycache=metadata', '-o', 'checksum=off', '-o', 'com.sun:auto-snapshot=false', 'antergos_hom2/home']: cannot create 'antergos_hom2/home': out of space"
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
** [Bugsnag] Notifying https://notify.bugsnag.com of exception
2016-02-08 13:18:59 [INFO] connectionpool.py(207) _new_conn(): Starting new HTTP connection (1): build.antergos.com
2016-02-08 13:18:59 [DEBUG] connectionpool.py(386) _make_request(): "GET /hook?cnchi=2580656b5c7be34e4144a3129a95e421&install_id=6843&result=False HTTP/1.1" 200 None
python3: no process found

[lots0logs]

@satriani-vai If you have time could you test the latest ISO with latest cnchi-dev package? I think I identified and fixed the issue. I was able to make two successful zfs installs earlier today.

[satriani-vai]

https://github.com/Antergos/Cnchi/issues/536 But note that I have set a "Set the home in a different home partition/volume" and chosen GPT partition scheme. I'll try without this feature and let you know. btw, for ZFS it would be called dataset :)

[satriani-vai]

Alright, finally it works, but without different home partition/volume. It's very simple zfs create $(zfs list -H -o name /)/home When the time comes with ZFS + LUKS? :)

EDIT: Take a look to this nice instruction: "Kick-ass (U)EFI+LUKS+ZFS booting" https://github.com/nathan7/bootr

EDIT2: It boots without problem but I see this one while booting http://i.imgur.com/hyIhILP.png

[karasu]

I think that the best would be to not tell zfs a predefined home size. I'll change it asap.

[satriani-vai]

I absolutely agree with you. That's the nice thing about ZFS. You don't need to define a dataset size. I love ZFS, I working about many years with ZFS on FreeBSD and have very good experience. I'm very happy, you have chosen this nice filesystem for Antergos. This makes Antergos even more interesting ;)

[nitrogensky]

Installed on a 30G VM and it set the ZFS Pool to 29.5G, but sudo zpool status and sudo zpool iostat -v show all is well. No errors or warnings.

[karasu]

@nitrogensky : If you can, could you post here a zpool status and a zfs list ?

[nitrogensky]

@karasu zpool status.txt zfs list.txt

I don't know what I am doing wrong, but I couldn't get the output to format corectly, so I hope these files are OK. Here's a gparted screenshot as well.

[karasu]

@nitrogensky : To avoid problems when booting, Cnchi creates a specific /boot partition with ext4. I'm not really sure this is necessary with the latest Grub2, though. That's why you get a 29,5GB pool, because 512MB are used by /boot

[karasu]

For future reference: https://github.com/Antergos/Cnchi/commit/d0e393b7cb64c8b2d1919191237a18479b0d3ff4

[itys]

This appears to continue to be an issue when installing ZFS over an existing partition scheme.

Current error number 236168, from antergos-18.1-x86_64.iso, Cnchi version 0.14.410 identified by the installer-update as current. "error wipefs probing initialization failed" "Cannot wipe the filesystem of device /dev/sda"

Effective workaround/fix is to reboot and restart the install.

As this has been open for a couple years, perhaps amend the error message to include "Please try again.. no, seriously, turn the system off and restart the install, it'll probably work the second time." ? :-)

[mrcasquinhas]

"'Cannot wipe the filesystem of device /dev/sda:wipefs: error /dev/sda: probing initialization failed: Device or resource busy' Please reference the following number when reporting this error: 965490"

This is the error that I got when choosing ZFS. So I started gparted, removed the 'unknown' partition and it is now unallocated. It told me to reboot, so I did. Installed again, with ZFS option and GPT, not MBR. All other options are at defaults. It works now, it's installing but it should go to the end without anymore issues. So the problem was partition was 'crazy' and unknow, so ZFS did not like it.