On Mon, 8 Apr 2019 at 10:33, Iain Buclaw <ibuclaw@xxxxxxxxxx> wrote: > > On Mon, 8 Apr 2019 at 05:01, Matt Benjamin <mbenjami@xxxxxxxxxx> wrote: > > > > Hi Christian, > > > > Dynamic bucket-index sharding for multi-site setups is being worked > > on, and will land in the N release cycle. > > > > What about removing orphaned shards on the master? Is the existing > tools able to work with that? > > On the secondaries, it is no problem to proxy_pass all requests to the > master whilst all rgw pools are destroyed and recreated. > > I would have though however that manually removing the known orphaned > indexes be safe though, side-stepping the annoying job of having to > force degrade the working service. > So is there no workaround for the bombarding of 404 requests on the sync master? I can see object entries in rgw.bucket.index and rgw.log, is there anywhere else it gets the list of things to sync? In somewhere encoded in omap? -- Iain Buclaw *(p < e ? p++ : p) = (c & 0x0f) + '0'; _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |