Ah. So that warning compares the objects per pg in that pool vs the objects per pg in the entire system, and if there is too much of a skew, issues a warning. If you look at 'ceph health detail' you will see some of the detail there. The reason you're seeing this is because you have lots and lots of pgs for the rgw pools that typically have almost no objects (.rgw.control, .rgw.root, .rgw.gc, .users, .users.email, possibly others). Generally speaking, it is only the .rgw.buckets pool that needs to get large (that is where most of the data goes). The pg_num value for others can be turned up as needed (see 'ceph df' to see where the objects are). It's not possible to decrease pg_num on a pool right now. For some of these pools, though, it is possible to create a new pool, copy the data, remove the old one, and rename the new one (while the radosgw daemons are stopped). The 'rados' tool has a copy function that can handle this, IIRC. You can also adjust the max skew to make this warning go away and change nothing (mon pg warn max object skew), but the large number of PGs will have a somewhat negative impact on cluster performance... sage On Thu, 5 Dec 2013, Alexis G?NST HORN wrote: > Hello, > > Here it is : > http://pastie.org/private/u5yut673fv6csobuvain9g > > > Thanks a lot for your help > > Best Regards - Cordialement > > Alexis G?NST HORN, > Tel : 0826.206.307 (poste ) > Fax : +33.1.83.62.92.89 > > > > IMPORTANT: The information contained in this message may be privileged > and confidential and protected from disclosure. If the reader of this > message is not the intended recipient, or an employee or agent > responsible for delivering this message to the intended recipient, you > are hereby notified that any dissemination, distribution or copying of > this communication is strictly prohibited. If you have received this > communication in error, please notify us immediately by replying to > the message and deleting it from your computer. > > > > 2013/12/4 Sage Weil <sage@xxxxxxxxxxxx>: > > Hi, > > > > Can you attach the output from 'ceph osd dump'? > > > > Thanks! > > sage > > > > > > > > "Alexis G?NST HORN" <alexis.gunsthorn@xxxxxxxxxxxx> wrote: > >> > >> Hello, > >> > >> I can't understand an error I have since now : > >> > >> HEALTH_WARN pool .rgw.buckets has too few pgs. > >> Do you have any ideas ? > >> > >> Some info : > >> > >> [root@admin ~]# ceph --version > >> ceph version 0.72.1 (4d923861868f6a15dcb33fef7f50f674997322de) > >> > >> [root@admin ~]# ceph osd pool get .rgw.buckets pgp_num > >> pgp_num: 10050 > >> > >> [root@admin ~]# ceph osd pool get .rgw.buckets pg_num > >> pg_num: 10050 > >> > >> [root@admin ~]# ceph -s > >> (...) > >> osdmap e30632: 201 osds: 201 up, 201 in > >> pgmap v4984359: 90666 pgs, 13 pools, 1276 GB data, 340 kobjects > >> 3916 GB used, 727 TB / 731 TB avail > >> 90666 active+clean > >> > >> > >> Thanks a lot, > >> Alexis > >> ________________________________ > >> > >> ceph-users mailing list > >> ceph-users@xxxxxxxxxxxxxx > >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > >> > > > > -- > > Sent from Kaiten Mail. Please excuse my brevity. > _______________________________________________ > 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
 |