For Ceph,this is fortunately not a major issue. Drives failing is considered entirely normal, and Ceph will automatically rebuild your data from redundancy onto a new replacement drive.If You're able to predict the imminent failure of a drive, adding a new drive /OSD will automatically start flowing data onto that drive immediately, thus reducing the time period with decreased redundancy.If You're running with very tight levels of redundancy, You're better off, creating a new OSD on a replacement drive before destroying the old OSD on the failed drive. But if You're running with anything near the recommended/default levels of redundancy, it doesn't really matter in which order you do it. Best regards, Simon Kepp, Kepp Technologies. On Tue, Dec 8, 2020 at 8:59 PM Konstantin Shalygin <k0ste@k0ste.ru> wrote:

Destroy this OSD, replace disk, deploy OSD.

k

Sent from my iPhone

...

On 8 Dec 2020, at 15:13, huxiaoyu@horebdata.cn wrote:

Hi, dear cephers,

On one ceph i have a failing disk, whose SMART information signals an impending failure but still availble for reads and writes. I am setting up a new disk on the same node to replace it. What is the best procedure to migrate data (or COPY ) from the failing OSD to the new one?

Is there any stardard method to copy the OSD from one to another?

best regards,

samuel

huxiaoyu@horebdata.cn _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io

---

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