[DRBD-user] LVM logical volume create failed

[Lars Ellenberg]

On Mon, Aug 27, 2018 at 08:21:35AM +0000, Jaco van Niekerk wrote:
> Hi
> 
> cat /proc/drbd
> version: 8.4.11-1 (api:1/proto:86-101)
> GIT-hash: 66145a308421e9c124ec391a7848ac20203bb03c build by mockbuild@, 2018-04-26 12:10:42
>  0: cs:SyncSource ro:Primary/Secondary ds:UpToDate/Inconsistent C r-----
>     ns:46466832 nr:0 dw:20 dr:46471876 al:1 bm:0 lo:0 pe:18 ua:0 ap:0 ep:1 wo:f oos:718000288
>     [>...................] sync'ed:  6.1% (701172/746540)M
>     finish: 5:54:53 speed: 33,700 (28,764) K/sec
> 
> lvcreate --addtag pacemaker --name lv_iscsi_lun0_snap --snapshot --size 100G /dev/vg_iscsi/lv_iscsi_lun0 --config 'activation { volume_list = [ "@pacemaker" ]}'
>   device-mapper: reload ioctl on  (253:3) failed: Device or resource busy
>   Aborting. Failed to activate snapshot exception store.
> 
> Is it because DRBD is still syncing that I can't process this command?

No.

I don't think this is related to DRBD.

You could add some -v to your lvcreate,
and look into the system and kernel logs for more information.
I suspect your lvm conf may be missing some filters,
or is otherwise not what you think it should be.

Or you are hitting a bug in lvcreate, unrelated to DRBD.
I seem to remember there used to be symptoms like this a while ago,
where not using lvmetad (use_lvmetad=0) was a workaround.
Something about how auto-activation worked, a race with udev,
and resulting attempt of "double-activation", which then failed.
Please add much hand-waving to the above, it's all somewhat vague
in the back of my head and I may be misremembering,
or confusing or mixing up multiple issues.

    Lars