On Tue, May 21, 2019 at 2:36 AM Strahil Nikolov <hunter86_bg@yahoo.com> wrote:

Hey Sahina,

it seems that almost all of my devices are locked - just like Fred's.

What exactly does it mean - I don't have any issues with my bricks/storage domains.

If the devices show up as locked - it means the disk cannot be used to create a brick. This is when the disk either already has a filesystem or is in use.

But if the device is a clean device and it still shows up as locked - this could be a bug with how python-blivet/ vdsm reads this

The code to check is implemented as

_canCreateBrick(device):
    if not device or device.kids > 0 or device.format.type or \
       hasattr(device.format, 'mountpoint') or \
       device.type in ['cdrom', 'lvmvg', 'lvmthinpool', 'lvmlv', 'lvmthinlv']:
        return False
    return True

Best Regards,

Strahil Nikolov

В понеделник, 20 май 2019 г., 14:56:11 ч. Гринуич+3, Sahina Bose <sabose@redhat.com> написа:

To scale existing volumes - you need to add bricks and run rebalance on the gluster volume so that data is correctly redistributed as Alex mentioned.

We do support expanding existing volumes as the bug https://bugzilla.redhat.com/show_bug.cgi?id=1471031 has been fixed

As to procedure to expand volumes:

1. Create bricks from UI - select Host -> Storage Device -> Storage device. Click on "Create Brick"

If the device is shown as locked, make sure there's no signature on device.  If multipath entries have been created for local devices, you can blacklist those devices in multipath.conf and restart multipath.

(If you see device as locked even after you do this -please report back).

2. Expand volume using Volume -> Bricks -> Add Bricks, and select the 3 bricks created in previous step

3. Run Rebalance on the volume. Volume -> Rebalance.

On Thu, May 16, 2019 at 2:48 PM Fred Rolland <frolland@redhat.com> wrote:

Sahina,

Can someone from your team review the steps done by Adrian?

Thanks,

Freddy