the main problem is not the pg_num,but some other problem about your network or ceph service AFAIK.
can your parse ceph -s ,ceph osd tree, cat ceph.conf ? 2016-05-23 11:52 GMT+08:00 Albert Archer <albertarche...@gmail.com>: > So, there is no solution at all ? > > On Sun, May 22, 2016 at 7:01 PM, Albert Archer <albertarche...@gmail.com> > wrote: > >> Hello All. >> >> Determining the number of pgs and pgps is the very hard job (of course >> for newbies like me ). >> The problem is, when we set the number of pgs and pgps for creating a >> pool, it seems there is no way to decrease the pgs for that pool. >> >> I configured 9 OSDs hosts (virtual machine in VMware ESXI ), in UP and IN >> state. and approximately 1700 pgs for rbd and another pool . >> >> but : >> ~1536 stale+active+clean >> ~250 active + clean >> >> so, how can i remove some pgs ???? or come back to ~1700 active+clean >> state ??? >> >> what is the problem ??? >> >> Regards >> Albert >> > > > _______________________________________________ > 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