Hello.
I've created an rgw installation, had uploaded about 60M files into a
single bucket. Removal had looked as a long adventure, so I "ceph osd
pool rm'ed" both default.rgw.data and default.rgw.index.
Now I have this:
# rados lspools
.rgw.root
default.rgw.control
default.rgw.meta
default.rgw.log
(same as for ceph osd pool ls)
but ceph -s shows:
pools: 6 pools, 256 pgs
Moreover, ceph osd df shows that I have TOTAL (raw) 5.5 TiB (use) 3.6
TiB (data) 3.4 TiB (omap) 35 GiB (meta) 86 GiB (avail) 1.9 TiB (%use)
65.36
I tried to force deepscrub for all OSDs but this didn't helped.
Currently I have a few tiny bits in all other pools and I don't
understand where the space is.
Installation is fresh nautilus, bluestore over HDD.
Few questions:
1. How this space is called? Lost? Non-gc? Cached?
2. Is it normal to have different number is lspools and total number of
pools?
3. Where I can continue to debug this?
4. (of course) how to this this?
Thanks!
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
