And if you put a big file in CephFS and then deleted it, the data will be deleted from the RADOS cluster asynchronously in the background (by the MDS), so it can take a while to actually get removed. :) If this wasn't the behavior then a file delete would require you to wait for each of those (10GB/4MB=)2560 objects to get deleted in separate network calls before returning! Software Engineer #42 @ http://inktank.com | http://ceph.com
On Wed, Apr 3, 2013 at 9:47 AM, John Wilkins <john.wilk...@inktank.com> wrote: > Can you elaborate on "manually deleted"? If you used an interface like RBD > or REST to upload the file, and then just deleted the upload from the file > system directly, your cluster map doesn't update. So you'd have a lost > object. > > > On Tue, Apr 2, 2013 at 2:35 AM, Adam Iwanowski <a.iwanow...@ogicom.pl> > wrote: >> >> Hello. >> >> Today i upgraded my cluster to version 0.60 and i noticed strange thing. >> I mounted to cepfs using kernel module, uploaded 10G file to test upload >> speed and then manually deleted file. From mountpoint i have no data in >> cluster but "ceph -w" still shows 10G data. Any idea how to clear that >> non-existing data from cluster, or is it bug in new version? >> >> Regards, >> Adam >> _______________________________________________ >> ceph-users mailing list >> ceph-users@lists.ceph.com >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > > > > > -- > John Wilkins > Senior Technical Writer > Intank > john.wilk...@inktank.com > (415) 425-9599 > http://inktank.com > > _______________________________________________ > ceph-users mailing list > ceph-users@lists.ceph.com > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
