On Tue, Oct 11, 2016 at 1:46 PM, Lu Dillon <ludi_1981@xxxxxxxxxxx> wrote: > Hi, John > > Thanks. Is this a new future of jewel? Now, we are using hammer. The path-based auth caps and restricting by namespace are new in Jewel, the pool layouts have existed much longer. I do not recommend using CephFS on anything older than Jewel. John > > > 发送自我的三星 Galaxy 智能手机。 > > > -------- 原始信息 -------- > 由： John Spray <jspray@xxxxxxxxxx> > 日期: 2016/10/11 19:17 (GMT+08:00) > 收件人： 卢 迪 <ludi_1981@xxxxxxxxxxx> > 抄送： ceph-users@xxxxxxxxxxxxxx > 主题： Re: can I create multiple pools for cephfs > > On Tue, Oct 11, 2016 at 12:05 PM, 卢 迪 <ludi_1981@xxxxxxxxxxx> wrote: >> Hi All, >> >> >> When I create meta and data pools for CEPHFS, all clients use the same >> pools >> for cephfs. Can I create multiple pools for different user? >> >> >> For example, I have pool_A for client A to mount cephfs; I want to create >> a >> pool_B for client B to mount cephfs. The purpose is to make sure client A >> can't see client B's files when they access the same cluster. >> >> Is this possible? Thanks. > > Yes, you can accomplish this. > > [the links below are offline right now but they should be accessible > as soon as ceph.com is back online] > > You can add multiple data pools to a filesystem (see add_data_pool > command) and then assign different pools to different directories > using layouts (http://docs.ceph.com/docs/jewel/cephfs/file-layouts) > > Then you can create client capabilities that restrict the clients by > path to a particular directory, and also restrict them to the pool > that is used for the data in that directory > (http://docs.ceph.com/docs/jewel/cephfs/client-auth/) > > You can also do this with rados namespaces instead of pools, if you > just want the security isolation without the overhead of having > entirely separate pools. > > John > > > >> >> >> _______________________________________________ >> ceph-users mailing list >> ceph-users@xxxxxxxxxxxxxx >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >> _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |