Hi Ketor, Can you tell what is making the ceph command fail at this point? If the mons are started before the OSDs, the crush map update ought to succeed... Thanks! sage On Mon, 11 Nov 2013, Ketor D wrote: > Hi Sage: > We always use mkcephfs to create out ceph cluster. Before > 0.72, all is ok. > But in 0.72, we found the osds can't start up useing "service ceph -a > start". After viewing the init-ceph.in script , I found that a return > value errro caused osds can't startup.And if we modifiy a little code > like 0.67.4 do, the osds can startup. > > Regards! > > Here is the patch: > > Signed-off-by: Ketor D <d.ketor@xxxxxxxxx> > --- > diff --git "a/src/init-ceph.in" "b/src/init-ceph.in" > index 7399abb..cf2eaa6 100644 > --- "a/src/init-ceph.in" > +++ "b/src/init-ceph.in" > @@ -331,7 +331,8 @@ for name in $what; do > -- \ > $id \ > ${osd_weight:-${defaultweight:-1}} \ > - $osd_location" > + $osd_location \ > + || :" > fi > fi > -- > 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 > > -- 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