Hello,
I'm testing in a lab test cluster all the upgrade procedures from an
old cluster (deployed originally as Jewel, now Luminous, osds are
already bluestore), but I'm stuck with a very odd observation (about
usage):
Luminous:
data:
objects: 27.88k objects, 109GiB
usage: 347GiB used, 11.1TiB / 11.4TiB avail
(after upgrade to) Mimic:
data:
objects: 43.24 k objects, 169 GiB
usage: 12 TiB used, 11 TiB / 22 TiB avail
Luminous:
GLOBAL:
SIZE AVAIL RAW USED %RAW USED
11.4TiB 11.1TiB 347GiB 2.97
POOLS:
NAME ID USED %USED MAX AVAIL OBJECTS
hdd-pool 1 48.5GiB 1.38 3.39TiB 12492
ssd-pool 2 60.0GiB 37.59 99.7GiB 15392
(after upgrade to) Mimic:
GLOBAL:
SIZE AVAIL RAW USED %RAW USED
22 TiB 11 TiB 12 TiB 51.41
POOLS:
NAME ID USED %USED MAX AVAIL OBJECTS
hdd-pool 1 108 GiB 3.26 3.1 TiB 27848
ssd-pool 2 60 GiB 38.96 94 GiB 15394
Can someone explains those diferences? (or what I missed here)
Is something related to bluestore_min_alloc_size?
Thanks,
--
Herbert
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
