Running trim / discard on an OSD

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]



Is there a way to run trim / discard on an OSD?

Long story:

I have a Proxmox-Ceph cluster with some OSD as storage for VMs. Discard works perfectly in this cluster. For lab and testing purposes I deploy Proxmox-Ceph clusters as Proxmox VMs in this cluster using nested virtualizacion, each with a few disks acting as OSD. Then a few VMs are configured in the nested Proxmox cluster using the nested Ceph as storage. Hope I've explained myself.

The problem I'm facing is that the nested Ceph OSD's are not sending the trim/discard command to the upstream Ceph OSDs, so thin provisioning is not kept. Somehow the discard chain is broken at some point.

I've tried using this on the nested ceph OSD:

ceph config set global bdev_enable_discard true
ceph config set global bdev_async_discard true

That somewhat helps, but it does not discard all the used space when data is deleted from the VM or the VM is fully removed, it maybe recovers like 20%. This is why I wonder if there is a way to run trim / discard in an OSD.

Thanks in advance

ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx

[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]

  Powered by Linux