On Thu, 15 Oct 2015, Tomy Cheru wrote: > Tried to write 160k objects of 64k size, totaling 10G to newstore via rgw > > PUTS were successful > > # ./rados df > .log 0 127 0 254 0 762 635 508 0 > .rgw 8 40 0 80 0 0 0 100 40 > .rgw.buckets 10485760 163840 0 327680 0 0 0 819200 10485760 > .rgw.buckets.index 0 20 0 40 0 655420 655400 655380 0 > .rgw.control 0 8 0 16 0 0 0 0 0 > .rgw.gc 0 32 0 64 0 96 64 64 0 > .rgw.root 1 3 0 6 0 3 2 5 5 > .users 1 5 0 10 0 0 0 5 5 > .users.email 1 3 0 6 0 0 0 3 3 > .users.swift 1 2 0 4 0 0 0 2 2 > .users.uid 2 6 0 12 0 134 131 136 8 > rbd 0 0 0 0 0 0 0 0 0 > total used 644956 164086 > total avail 194620640 > total space 195265596 > > however not able to see at least 10G worth data in device > > # du -ks dev/osd0/ > 611548 dev/osd0/ > # du -ks dev/osd0/db/ > 611524 dev/osd0/db/ > # du -ks dev/osd0/fragments/ > 0 dev/osd0/fragments/ > > why this discrepancy ?? > > Does newstore apply any kind of storage efficiency in io path ?? In current master rocksdb has snappy compression enabled by default, and I'm guessing you're writing either 0's or something very compressible. If you try wip-newstore-frags cmopression is disabled (it's problematic in general)... sage -- 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
