Bug #23554
openmon: mons need to be aware of VDO statistics
0%
Description
I created an OSD on top of a logical volume with a VDO device underneath.
Ceph is unaware of how much compression and deduplication has been done.
[root@reesi003 ~]# ceph osd df | grep 'ID\|^55' ID CLASS WEIGHT REWEIGHT SIZE USE AVAIL %USE VAR PGS 55 hdd 3.65810 1.00000 3745G 55080M 3692G 1.44 0.06 367 [root@reesi004 ~]# vdostats --hu Device Size Used Available Use% Space saving% /dev/mapper/vdo_sda 3.6T 25.1G 3.6T 0% 21%
Note the space saving percentage is incorrect. I filed https://bugzilla.redhat.com/show_bug.cgi?id=1563794 for that.
Steps to reproduce:
vdo create --name vdo_sda --device=/dev/sda --writePolicy=async pvcreate /dev/mapper/vdo_sda vgcreate vg_sda /dev/mapper/vdo_sda lvcreate -l 100%FREE -n lv_sda vg_sda ceph-volume lvm create --bluestore --data vg_sda/lv_sda --block.db journals/sda
Updated by Greg Farnum about 6 years ago
What would we expect this monitor awareness to look like? Extra columns duplicating the output of vdostats?
Updated by Josh Durgin about 6 years ago
At least this: https://github.com/ceph/ceph/pull/20516
Updated by David Galloway about 6 years ago
Of course Sage is already on it :)
I don't know where the osd df
output is read from but my non-developer mind assumes the OSD process could run vdostats
and report back how much actual space is being used.
Based on what I'm actually able to decipher in Sage's PR, that's what's being done.
Updated by Greg Farnum about 6 years ago
Right, but AFAICT the monitor is then not even aware of VDO being involved. Which seems fine to my naive thoughts, but might be an issue in terms of either "advertising" or accurate trend monitoring.
Updated by Patrick Donnelly about 6 years ago
- Subject changed from MONs need to be aware of VDO statistics to mon: mons need to be aware of VDO statistics
- Source set to Development
Updated by Patrick Donnelly about 6 years ago
- Project changed from Ceph to RADOS