HI Yehuda,
Thanks for the answer. We are planning to use the
CEPH_OSD_OP_DELETE to implement
fallocate/hole punching, the details are described in the blueprint
(wiki.ceph.com). Your comments/suggestions
are appreciated.
Regards,
Li Wang
On 04/29/2013 10:23 PM, Yehuda Sadeh wrote:
On Sun, Apr 28, 2013 at 8:43 AM, 汪黎 <liwang@xxxxxxxxxxxxxxx> wrote:
Hi Yehuda, We have a code question to consult. For unlink() system call to delete a file, does the Ceph kernel client go through ceph_unlink()? If so, it seems from the code tht client only communicate with MDS for metadata operation, client seems not contact with OSD to delete the object. So, is that MDS takeing the responsibility of deleting file data? And, we find the definition of CEPH_OSD_OP_DELETE (include\linux\ceph\rados.h), however, it seems it never actually used in the kernel client code, is it left for future? Thanks for your time. Regards, Li Wang
Lets keep this discussion on the mailing list so everyone can benefit
from the discussion, cc'ing it here.
As for your questions, the MDS is responsible for removing the
objects. The kernel fs client code has probably no need for that
operation.
Yehuda
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
