All of lore.kernel.org
 help / color / mirror / Atom feed
From: Robert White <rwhite@pobox.com>
To: Chris Murphy <lists@colorremedies.com>,
	Markus Moeller <huaraz@moeller.plus.com>
Cc: Btrfs BTRFS <linux-btrfs@vger.kernel.org>
Subject: Re: Btrfs subvolume question
Date: Wed, 04 Feb 2015 19:40:30 -0800	[thread overview]
Message-ID: <54D2E62E.7080208@pobox.com> (raw)
In-Reply-To: <CAJCQCtSXJZ4paKiZBNDOp-yKccrLbizmGDj_8y5k5EC-9DFjXg@mail.gmail.com>

On 02/04/2015 06:27 PM, Chris Murphy wrote:
> On Wed, Feb 4, 2015 at 3:54 PM, Markus Moeller <huaraz@moeller.plus.com> wrote:
>> Hi ,
>>
>>    I am new to btrfs and wonder what I need to do to move subvolumes to the
>> right filesystem.  I see the following:
>>
>> df -h
>> Filesystem                       Size  Used Avail Use% Mounted on
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /
>> devtmpfs                         235M  8.0K  235M   1% /dev
>> tmpfs                            242M   84K  242M   1% /dev/shm
>> tmpfs                            242M  2.4M  240M   1% /run
>> tmpfs                            242M     0  242M   0% /sys/fs/cgroup
>> /dev/mapper/system_13.2-usr_lv    18G  6.9G   10G  41% /usr
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /srv
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /.snapshots
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /tmp
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /usr/local
>> /dev/sda2                        486M   59M  398M  13% /boot
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32%
>> /boot/grub2/x86_64-efi
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /boot/grub2/i386-pc
>> /dev/mapper/export_vg-export_lv   20G   18G  2.9G  86% /export
>> /dev/mapper/export_vg-src_lv     5.0G  2.6G  2.5G  51% /src
>> /dev/mapper/system_13.2-var_lv   4.0G  196M  3.4G   6% /var
>> /dev/mapper/system_13.2-opt_lv   6.0G  152M  5.3G   3% /opt
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/spool
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/tmp
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/opt
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/log
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/lib/named
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/lib/mailman
>> /dev/mapper/system_13.2-root_lv  5.0G  1.5G  3.2G  32% /var/lib/pgsql
>>
>>
>> and I would like to have all the var subvolumes using the /var filesystem
>> space and not root.  The same for /boot/grub2 subvolumes. They should use
>> the /boot filesystem.
>
> The giveaway is in the first column. All of these are device mapper
> LVM LV's, they are not Btrfs subvolumes, in fact they're separate
> Btrfs volumes (unique, unrelated filesystems). What's suspicious,
> though, is that a bunch of these LV's have exactly 1.5G Used, which
> makes no sense. For example, /boot/grub2/i386-pc has 1.5G used, but
> that's not possible, the entire GRUB2 package is maybe 50MB.

Actually given that the first column repeats 
/dev/mapper/system_13.2-root_lv as the device, I think he's repeatedly 
mounted the same device filesystem with different subovol= directives.

That's why so many of them have the same 1.5G used.

In terms of actual LVs he only seems to have _root_ _opt_ _var_ and 
_usr_ once you strip away the various noise components.

In particular oddity fassion /var is a different device but /var/log 
goes back to the root device.

In a slightly over-thought way its kind of twisted genius if you view 
/var/lib/named and such as being tightly bound to /etc. (/var/tmp is 
totally wrong.)

It's not something I would want to maintain either.

They also might be bind mounts instead of subvol= mounts. I'd have to 
see /etc/fstab.

The output of btrfs subvol list  on / /var /usr and /opt would be nice 
as well.


>
> Honestly, I suggest reinstalling and don't deal with the tricky task
> of converting all of this. It'll take longer for me to explain, and
> you to read, how to fix this without reinstalling. I also suggest
> making sure that you don't put Btrfs on LVM, there's no good reason to
> do this for most use cases.
>
>
>   I don't know how I got to this and don't know how to
>> change.
>
> It's not your fault. This is a pathological side effect of openSUSE
> 13.2's, quite frankly bizarre, layout. By default it creates a single
> Btrfs volume, and makes all of these subvolumes, which is strange
> enough on its own because it's overly complicated (just look at
> /etc/fstab) for no good reason. But somehow in your case, you ended up
> with a bunch of LVM LVs first, and then each of those is a Btrfs
> volume. If it were me, I'd file a bug report against 13.2, and then
> file a feature request citing the bug report. If you don't want to do
> both, then just file the feature request against Factory because
> installer related bug reports on openSUSE basically get rejected
> seeing as there's nothing that can be done about it in Distribution
> now anyway.
>


  reply	other threads:[~2015-02-05  3:48 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-02-04 22:54 Btrfs subvolume question Markus Moeller
2015-02-05  2:27 ` Chris Murphy
2015-02-05  3:40   ` Robert White [this message]
2015-02-05  6:03     ` Chris Murphy
2015-02-05 20:28       ` Markus Moeller
2015-02-05 20:58         ` Chris Murphy
2015-02-05 22:49           ` Markus Moeller
2015-02-06  3:05             ` Chris Murphy
2015-02-06 22:29               ` Markus Moeller
2015-02-07  0:14                 ` Chris Murphy

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=54D2E62E.7080208@pobox.com \
    --to=rwhite@pobox.com \
    --cc=huaraz@moeller.plus.com \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=lists@colorremedies.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.